'oei'chcu/, Wm> nJ&gr? /> * CZ^O & THE MARINE ALGAE OF THE PACIFIC COAST OF NORTH AMERICA PART III MELANOPHYCEAE BY WILLIAM ALBERT SETCHELL AND NATHANIEL LYON GARDNER - u~ _ C~ o — — . m ^ ^— ~- CD University of California Publications in Botany m == ^ Vol. 8, Part III, pp. 383-898, plates 34-107 i— ^ m UNIVERSITY OF CALIFORNIA PRESS BERKELEY, CA S 1925 ^ BERKELEY, CALIFORNIA UNIVERSITY OF CALIFORNIA PUBLICATIONS Note. The University of California Publications are offered in exchange for the publi- cations of learned societies and institutions, universities and libraries. Complete lists of all the publications of the University -will be sent upon request. For sample copies, lists of publications, and other information, address the MANAGER OF THE UNIVERSITY PRESS, BERKELEY, CALIFORNIA, U. S. A All matter sent in exchange should be addressed to THE EXCHANGE DEPARTMENT, UNIVERSITY LIBRARY, BERKELEY, CALIFORNIA, U. S. A. Publications of the University of California Press may also be obtained from THE CAMBRIDGE UNIVERSITY PRESS, FETTER LANE, LONDON, E.C. 4, ENGLAND, to which orders originating in Great Britain and Ireland should be sent. BOTANY.— W. A. Setchell and R. M. Holman, Editors. Price per volume, $3.50 (voL 5, $6.00) ; beginning with volume 6, $5.00. Volumes I, n, m, IV, V, VI, VII, and IX completed. Volumes VIII, X, XI, XII, and XIII in progress. Cited as Univ. Calif. PubL Bot. Vol. 6. 1912-1922. 1. Studies in Nicotiana. L by William Albert SetchelL Pp. 1-86. December, 2. Quantitative Studies of Inheritance in Nicotiana Hybrids. I, by Thomas Harper Goodspeed. Pp. 87-168, plates 1-28. December, 1912 L00 8. Quantitative Studies of Inheritance in Nicotiana Hybrids. It, by Thomas Harper Goodspeed. Pp. 169-188, plates 29-84. January, 1918 .20 4. On the Partial Sterility of Nicotiana Hybrids iaade with N. sylvestris as a Parent, by Thomas Harper Goodspeed. Pp. 189-198. March, 1913 .10 5. Notes on the Germination of Tobacco Seed. I, by Thomas Harper Good- speed. Pp. 199-222. May, 1913 - -25 6. Quantitative Studies of Inheritance in Nicotiana Hybrids, in, by Thomas Harper Goodspeed. Pp. 223-231. April, 1915 .„ - 10 7. Notes on the Germination of Tobacco Seed. II, by Thomas Harper Good- speed. Pp. 283-248. June, 1915 . - «15 8. Parthenogenesis, Parthenocarpy and Phenospenny in Nicotiana, by Thomas Harper Goodspeed. Pp. 249-272, plate 35. July, 1915 .25 9. On the Partial Sterility of Nicotiana Hybrids made with N. sylvestris as a Parent. H, by T. H. Goodspeed and A. H. Ayres. Pp. 273-292, plate 36. 10. On the Partial Sterility of Nicotiana Hybrids made with N. sylvestris as a Parent, m, An Account of the Mode of Floral Abscission in the Fj Species Hybrids, by T. H Goodspeed and J. N. Kendall. Pp. 293-299. November, 11. The Nature of the F, Species Hybrids between Nicotiana sylvestris and Varieties of Nicotiana Tabacum, with Special Reference to the Conception of Reaction System Contrasts in Heredity, by T. H. Goodspeed and R. E. Clausen. Pp. 301-346, plates 37-48. January, 1917 .45 12. Abscission of Flowers and Fruits in the Solanaceae, with Special Reference to Nicotiana, by John N. Kendall. Pp. 347-428, 10 text figures, plates 49- 53. March, 1918 ~ • — - — «85 13. Controlled Pollination in Nicotiana, by Thomas Harper Goodspeed and Pirie Davidson. Pp. 429-434. August, 1918 - 10 14. An Apparatus for Flower Measurement, by T. H. Goodspeed and R. E. Clausen. Pp. 435-437, plate 54, 1 figure in text. September, 1918 „. .05 15. Note on the Effects of Illuminating Gas and Its Constituents in Causing Abscission of Flowers in Nicotiana and Citrus, by T. H. Goodspeed, J. M. McGee and R. W. Hodgson. Pp. 439-450. December, 1918 JL0 16. Notes on the Germination of Tobacco Seed, m, Note on the Relation of Light and Darkness to Germination, by T. Harper Goodspeed. Pp. 451- 455. April, 1919 .. - .05 17. Inheritance in Nicotiana Tabacum. LA Report on the Results of Crossing Certain Varieties, by William Albert Setchell, Thomas Harper Goodspeed, and Roy Elwood Clausen. Pp. 457-582, 2 figures in text, plates 55-85. April, 1922 _ - ~ — • 1.76 ' University of California Publications in BOTANY VOLUME VIII Part III 1925 WILLIAM ALBERT SETCHELL R. M. HOLMAN EDITORS UNIVERSITY OF CALIFORNIA PRESS BERKELEY, CALIFORNIA THE MARINE ALGAE OF THE PACIFIC COAST OF NORTH AMERICA PART III MELANOPHYCEAE BY WILLIAM ALBERT SETCHELL AND NATHANIEL LYON GARDNER UNIVERSITY OF CALIFORNIA PRESS BERKELEY, CALIFORNIA University of California Publications in Botany Vol. 8, Part III, pp. 383-898, plates 34-107 Issued June 26, 1925 The University of California Press Berkeley, California The Cambridge University Press London, England CONTENTS PAGE Subclass 3. Melanophyceae 387 Series 1. Phaeosporeae 389 Order 1. Sphacelariales 390 Family 1. Sphaeelariaceae 391 1. Sphacelaria 392 2. Chaetopteris 397 Order 2. Ectocarpales 398 Family 2. Ectocarpaceae 400 3. Pylaiella 401 4. Ectocarpus 407 5. Streblonema 440 Family 3. Myrionemataceae 453 6. Myrionema 454 7. Compsonema 473 8. Hecatonema 488 9. Microspongium 492 Family 4. Ralfsiaceae 493 10. Ralfsia 494 11. Hapterophycus 499 12. Lithoderma 500 Family 5. Elachisteaceae 502 13. Elachistea 503 14. Gonodia 505 Family 6. Leathesiaceae 507 15. Petrospongium 508 16. Leathesia 510 Family 7. Punctariaceae 513 17. Punctaria 515 Family 8. Asperococcaceae , 522 18. Halorhipis 523 19. Soranthera 525 20. Myelophycus 527 Family 9. Striariaceae 528 21. Striaria 529 22. Stictyosiphon 530 Family 10. Scytosiphonaceae 530 23. Scytosiphon 531 24. Ilea 535 25. Endarachne 538 26. Colpomenia 539 27. Hydroclathrus 542 Family 11. ^Egiraceae 543 28. iEgira 544 29. Meneghiniella 548 Family 12. Heterochordariaceae 549 30. Heterochordaria 550 [383] ::si University of California Publications in Botany t VoL - PAGE Family 13. Chnoosporaceae 552 31. Chnoospora 552 Order 3. Desmarcstiales 554 Family 14. Myriogloiaceae 555 32. Myriogloia 556 Family 15. Dcsmarestiaceae 558 33. Desmarestia 559 Order 4. Chordariales 570 Family 16. Chordariaceae 571 34. Chordaria 571 35. Analipus 575 36. Gobia 576 Family 17. Coilodesmaceae 577 37. Coilodesme 578 38. Phaeostrophion 585 Order 5. Dictyosiphonales •' 586 Family 18. Dictyosiphonaceae 587 39. Dictyosiphon 587 Order 6. Laminariales 590 Family 19. Chordaceae 591 40. Chorda 592 Family 20. Laminariaceae 593 Tribe 1. Laminarieae 594 41. Laminaria 594 Tribe 2. Cymathaereae 606 42. Pleurophycus 606 43. Cymathaere 608 Tribe 3. Agareae 609 44. Costaria 609 45. Thalassiophyllum 613 46. Agarum 614 Tribe 4. Hedophylleae 616 47. Hedophyllum 617 48. Arthrothamnus 619 Family 21. Lessoniaceae 621 Tribe 1. Lessonieae 621 49. Dictyoneurum 621 50. Nereocystis 623 51. Postelsia 624 Tribe 2. Macrocysteae 626 52. Macrocystis 626 53. Pelagophycus 629 Tribe 3. Lessoniopseae 631 54. Lessoniopsis 631 Family 22. Alariaceae 633 Tribe 1. Alarieae 633 55. Pterygophora 634 56. Alaria 635 Tribe 2. Ecklonieae 645 57. Eisenia 646 Tribe 3. Egregieae 647 58. Egregia 647 1!,2, l SeteheH-Gao'dner: Melanophyceae 385 PAGE Series 2. Aplanosporeae 649 Order 7. Dictyotales g 50 Family 23. Dictyotaceae 650 59. Dictyota 652 60. Neurocarpus 655 61. Taonia 656 62. Chlanidophora 658 63. Zonaria 659 64. Padina 661 Series 3. Cyclosporeae 662 Order 8. Fucales 663 Family 24. Fucaceae 663 65. Fucus 664 66. Pelvetia 7qq 67. Pelvetiopsis 702 68. Hesperophycus 703 Family 25. Sargassaceae 704 69. Blossevillea 705 70. Cystophyllum 706 71. Halidrys 707 72. Cystoseira 708 73. Sargnssum 72 j The third part of the Marine Algae of the Pacific Coast of North America, comprising an account of the Melanophyceae or Brown Algae, succeeding the accounts of the Myxophyceae and Ohloro- phyceae, is presented without introduction or explanation, pending the publication of the final part, which is in advanced preparation. It is intended to issue with the completed volumes a statement as to the principles followed, methods used, territory covered, sources of information, material, etc., as well as to make acknowledgment to the various authorities, students and collaborators who have been of the greatest assistance in carrying out the long and laborious task. W. A. Setchell and N. L. Gardner. THE MARINE ALGAE OF THE PACIFIC COAST OF NORTH AMERICA PART III BY WILLIAM ALBERT SETCHELL AND NATHANIEL LYON GARDNER Subclass 3. MELANOPHYCEAE stiz. Multicellular thallophytes containing- the green pigment, chloro- phyll (including xanthophyll), and the brown pigment, phycophaein ; thallus varying from simple or slightly branched filaments of a single row of cells and simple membranes varying from a single layer to several layers of similar cells, to solid plant bodies of varying form, e.g., membranous, crustaceous, globular, hemispherical, filamentous, etc.. and with differentiated internal and external tissues; cells with distinct cell walls, cytoplasm, at least the outer cells containing definite chromatophores, carrying the pigments, and typically uninucleate, although at times becoming multinucleate; assimilation products various carbohydrates and oils, but never true starch; internal cells in more complex forms elongated and forming definite mechanical conducting and storage systems as distinct from the outer tissues of absorbing and assimilating cells ; multiplication by splitting and by gemmae of more or less distinctive form ; reproduction by both motile and non-motile cells, motile cells with one or two lateral, or at least subapical, cilia ; non-sexual reproduction by 2-ciliated zoospores or by aplanospores (Dictyotales) : sexual reproduction by the fusion of two equal, or of two unequal, motile gametes, or by the fusion of a smaller motile 2-ciliated sperm and a larger non-motile egg (Lamina- riales and Fucales) or by the fusion of a smaller 1-ciliated sperm and a larger non-motile egg (Dictyotales) ; an antithetic alternation of generations present in many, if not all, species; gametophytes usually unisexual, similar to, up to extremely different from, the sporophytes. Melanophyceae Stizenberger, Dr. Ludwig Rabenhorst's Algen Sachsens, 1860, p. 36; Rabenhorst, Flora Eur. Alg., vol. 3, 1868, p. 393. Fucoideae Agardh, Syn. Alg. Scand., 1817, p. IX. Melama- spermeae Harvey, in Mackay, Flora Hibern., part III. 1836, p. 157 : 3SS University of California Publications in Botany [Vol.8 Gen. So. African Plants, 1838, p. 393. Phaeophyceae Kjellman, in Engler and Prantl, Die natiirl. Pflanzenfam., Teil 1, Abt. 2 (Lief. 60), 1891, p. 176. The Melanophyceae, or Brown Algae, are, with very few exceptions, marine and are, properly, all multicellular. We feel convinced that the termination -phyceae should be retained for the subclasses of the class Algae, or Phycophyta. Since Stizenberger's Melanophyceae, 1860, applies to the same group of plants as does Kjellman 's Phaeo- phyceae, 1891, and since Stizenberger's term seems equally applicable, we have chosen it in preference to Kjellman 's, even though Phaeo- phyceae is at present in general use. The Melanophyceae are much more complex in both structure and reproductive processes than are the Chlorophyceae or the Myxophyceae. In size, they range from minute disks to plants of considerable size, reaching among the kelps to a length of possibly 100 meters or longer. It has seemed the more natural to restrict the Melanophyceae to the multicellular forms than to include the unicellular forms with chlorophyll and a brown pigment. The Bacillariales, or Diatoms, and the Peridi.niales, as well as certain groups of the Flagellates, even if they may be regarded as belonging to the vegetable kingdom, seem to have no characters that would associate them with the Melanophyceae except the possession of a similar brown pigment. The Melanophyceae are, with few exceptions, viz., Phacosaccion Farlow and Omphalophylhon Rosenvinge, provided with specialized zoosporangia or with gametangia or with both. In one order, viz., Dictyotales, the non-sexual spores are always motionless. There is an antithetic alternation of generations in most of the groups of the Melanophyceae, the Fucales being the only order in which it is certainly absent, where the sporophyte generation of this order is considered as represented by the products of the division of the oogonial cell. In the Sphacelariales and the Ectocarpales it is not so clearly proved as yet, but Kylin (1918) has demonstrated reduction divisions of the primary nucleus of the zoosporangium in certain species. The two generations in both of these orders are alike in size. We find species, however, in which the gametophyte is microscopic, in striking contrast to the large, often even gigantic, sporophyte. The same relation between gametophyte and sporophyte is suspected, but not yet observed, in the Desmarestiales, Chordariales, and Dictyosi- phonales. In the Cutleriales the two generations may be similar or dissimilar, while in the Dictyotales they are similar in size and vegetative structure. 19-5] Setchett-Gardner: Melanophyceae 389 The classification of the Melanophyceae is in a very unsatisfactory condition, that of Kjellman (1891) and that of Oltmanns (1904) seeming very imperfect, especially since the later studies of Sauvageau, Kylin, and others have changed our views in so many important respects. The more recent classifications of Kylin (1917), of Oltmanns (1922) and of William Randolph Taylor (1922) have assisted greatly toward a more natural arrangement. We have therefore adopted many of the views of Kylin (1917) and of Oltmanns (1922) and have incor- porated some additional ideas of our own. Oltmanns has divided the subclass into seven orders. In this, we are inclined to follow him, but think it better to regard some of the other divisions also as of equal rank with these. According to our scheme, the Melanophyceae of the Pacific Coast of North America are arranged in the following manner : Key to the Series 1. Reproduction sexual only 3. Cyclosporeae (p. 662) 1. Reproduction both sexual and non-sexual 2 2. Aplanospores present 2. Aplanosporeae (p. 649) 2. Aplanospores absent 1. Phaeosporeae (p. 389) Series 1. PHAEOSPOREAE thuret Thallus composed of simple or branching filaments either of a single row of cells (monosiphonous) or solid with diverse tissues or of simple or solid membranes, or cushion form, branched or unbranched, sometimes hollow, variously aggregated; non-sexual reproduction (where known) by zoospores with 2 lateral cilia produced in unilo- cular zoosporangia ; aplanospores never present ; gametangia pluri- locular. or in one order (Laminariales) unilocular, male gametes motile, biciliated, female gametes motile and biciliated or non-motile; gametophyte and sporophyte similar or more or less dissimilar. Phaeosporeae Thuret, Rech. sur les zoosp. I, 1850, p. 233 (p. 24. repr.) ; Kjellman, in Engler and Prantl, Xatiirl. Pflanzenfam., Teil 1. Abt. 2 (Lief. 60), 1891, p. 180 (in part) ; Oltmanns, Morph. und Biol, der Algen, vol. 1, 1904, p. 348. Phaeozoosporeae De-Toni, Syst. Uebers. Fueoid., 1891, p. 175. Phaeosporales Oltmanns, Joe. cit., p. 348. The Phaeosporeae of Thuret have generally been regarded as con- stituting an order, but they seem to us rather to represent a series under the subclass Melanophyceae composed of the orders Sphace- lariales, Ectocarpales, Cutleriales, Dictyosiphonales, Desmarestiales, and possibly also the Laminariales. The general characteristic is the occurrence of the typical brown zoospores of the Melanophyceae. These are, as yet, unknown in some genera and species of the series, while r~ 390 University of California Publications in Botany I VoL - s in others they occur hut seldom. What our increasing' knowledge may teach us concerning these seemingly exceptional forms must be Left for future consideration. The typical members of the Phaeosporeae possess unilocular zoosporangia and plurilocular gametangia and have the garnet ophyte and sporophyte of very closely similar size and struc- ture. Some scattering genera and species are only known to produce the unilocular zoosporangia, while others produce only the plurilocular gametangia, but, considering seemingly close affinities in these puzzling cases, it seems undesirable for the present, at least, to refer either elsewhere than to the Phaeosporeae. They may be found to repre- sent either imperfect, or perhaps even dimorphic forms, when their development shall have been carefully studied. The Laminariales, which we still include under the Phaeosporeae, were long known only with unilocular sporangia producing biciliated zoospores, but have recently been found also to possess a microscopic gametophyte with decidedly modified (unilocular) gametangia producing either non- motile egg cells (or oospheres) or biciliated sperms. The Phaeosporeae may be distinguished from the Acinetosporeae by the lack of aplano- spores, and from the Fucales by the possession of either known zoo- sporangia or plurilocular gametangia or both. Key to the Orders 1. Gametangia present on macroscopic plants (zoosporangia known or unknown).... 2 1. Only zoosporangia present on macroscopic plants (gametangia known, or suspected of being, on microscopic plants) 3 2. Terminal cell present and conspicuous 1. Sphacelariales (p. 390) 2. Terminal cell, if present, at least not conspicuous. 2. Ectocarpales (p. 398) 3. Growth of zoosporangial frond, strictly terminal 5. Dictyosiphonales (p. 586) 3. Growth of zoosporangial frond not strictly terminal 4 4. Growth of zoosporangial frond trichothallic 3. Desmarestiales (p. 554) 4. Growth of zoosporangial frond not trichothallic 5 5. Growth apical, from subapical cells 4. Chordariales (p. 570) 5. Growth never apical, either intermediate or near the base .. 6. Laminariales (p 590 ) Order 1. SPHACELARIALES oltmanns Filamentous, generally tufted and branched Phaeosporeae, rarely monosiphonous throughout, usually monosiphonous near or at the tips, becoming polysiphonous below through longitudinal walls, and in some cases corticated slightly or considerably by descending filaments, attached by horizontal rhizoids often cohering to form a disk, or, in some cases, arising early and forming a more or less complex hori- zontal thallus giving rise to the erect filaments ; growth of each erect 1925] SetcheU Gardner: Melanophyceat 391 filament from a more or less conspicuous apical cell; older cell Avails becoming dark colored when treated with eau de javelle; chromato- phores numerous in each cell, usually disk-shaped; zoosporangia uni- locular, large ; gametangia pluriloeular, similar or with some distinc- tion as to size and number of loculi (mega- . 3. Ectocarpus acutus S. and G. Fronds 5-9 cm. high, dark brown when dried, olive green when living, feathery; erect filaments more or less entangled and fasciculate below, free above, profusely branched, main filaments densely cor- ticated; branches mostly alternate, strict; ultimate ramuli often 41<) University of California Publications in Botany [Vol.8 secund, acute-subulate ; cells slightly doliiform, 40-60/x diam., 0.4-2 times as long as broad (usually shorter than the breadth) ; ehromato- phores thick, irregularly branched bands, few in a cell, containing several pyrenoids ; zoosporangia unknown ; gametaugia numerous and regularly placed, cylindrico-conical, 100-150/* (up to 230//.) long, 20-35/a broad. Growing, for the most part at least, on larger Melanophyceae. Pugel Sound to central California (Carmel). Setchell and Gardner, Phyc. Cont. VI, 1922, p. 404. pi. 48, figs. 36-39 and pi. 49, figs. 40, 41. Ectocarpus penicillatus Saunders. Phyc. Mem., 1898, p. 155, pi. 21, figs. 3, 4 (not of Kjellman). Ectocarpus eonfervoides f. acumiyiatus Collins, in Setchell and Gardner, Alg. N.W. Amer, 1903, p. 237, Mar. Alg. Vancouver Is., 1913, p. 106; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), nos. 1033, 1127. The type of this species is no. 2886 of Gardner, collected at Carmel. California, in May, 1915, and growing on Desmarestia herbaeea. It seems to be the same as the plant figured (and described) by Saunders (Joe. cit.) under Ectocarpus penicillatus. It is undoubtedly the same plant as that described by F. S. Collins under the name E. eonfer- voides f. acuminatus which was founded on no. 235 of Gardner (Herb. Univ. Calif', no. 99022) collected at Whidbey Island, Washington, growing on a broad form of Desmarestia, Its affinities are closely with Ectocarpus eonfervoides f. ty picas, but it has shorter cells, sharper ramuli, decidedly constricted filaments, with the gametaugia more inclined toward cylindrical than is general in that form. The chro- matophores are more distinctly band-shaped and more separated from one another than in E. eonfervoides. A few small but profusely fruiting specimens of plants seemingly of this species have been found growing on Myriogloia Andersonii at Neah Bay, Washington (Gardner, no. 3817). These were attached to the host by means of penetrating rhizoidal filaments, particularly by the descending corticating filaments. The specimens differ from the typical form in being smaller, in having shorter and blunter game- tangia and in having on separate individuals seriate zoosporangia ( ?) some of which are divided longitudinally into four loculi. Further study of more material will be necessary to determine the status of this form. For the present we are placing it with E. acutus. The zoosporangia are very numerous, and they resemble in form those figured by Sauvageau (1896&, p. 33) for Ectocarpus virescens. They, 1925] Setchdl-Gardner: Melcmophyceae 417 however, have a glistening: appearance, as though they were abortive or in a pathological condition, while those represented by Sauvageau are producing zoospores. This condition in which we find ours is quite common among our Pacific Coast species of Ectocarpus. We have not yet seen a single specimen in which the so-called seriate zoosporangia show any indication of producing zoospores. 4. Ectocarpus tomentosus (Huds.) Lyngb. Fronds tufted or caespitose, consisting of masses of erect, simple or branched, rope-like strands, yellowish brown to dark brown ; pros- trate filaments short, irregularly branched ; erect filaments twisted together into rope-like masses, 1 mm. to 20 cm. high; main branches long, divergent ; ultimate ramuli profuse, alternate, irregularly spaced, never opposite, mostly short and patent, subpiliferous, often hooked at the tips; cells of the main erect filaments 8-12/* broad, 1.5-2.5 times as long as broad, especially below, mostly quadrate above ; ehromato- phores irregularly band-shaped, few in each cell ; zoosporangia ovoid, terminal or lateral, 28-36/* long, 20-26/x broad; gametangia lateral, patent, often secund, straight or recurved, sessile or on short pedicels, 25-75/* long, 10-15/t broad. Growing on members of the Fucaceae. Extending from Alaska (Harvester and Kadiak Islands) to southern California (Laguna). Lyngbye, Hydrophyt. Dan., 1819, p. 132, pi. 44A 1-3; Greville, Crypt, Flor., vol. 6, 1828, pi. 316; Saunders, Phyc. Mem., 1898, p. 155, pi. 24, figs. 1, 2, Alg. Harriman Exp., 1901, p. 417; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 238. Ectocarpus tomcntosoides Guernsey, Notes on Mar. Alg., 1912, p. 198. Conferva torn cut osa Hudson, Fl. Angl., 1762, p. 480 (in part). (Ed. 2, 1778, p. 594.) Our idea of Ectocarpus tomentosus dates back definitely only to Dillwyn. His" interpretation differs somewhat from that of botanists earlier than his time, but is probably as reliable an interpretation of the species of Hudson as may be possible. Much hinges on the ideas as to the identity of the plant of Dillenius (1741, p. 19, pi. 3, fig. 19) figured and described under the name of Conferva marina tome nt osa, minus tenera et ferruginea which is quoted by Hudson among the synonyms of his species. The figure of Dillenius is not reassuring and Dawson Turner (Trans. Linn. Soc, vol. 7, 1804. p. 105), as quoted by Dillwyn, states that the specimen in the Dillenian Her- barium "is only a bad specimen of C. littoral is." Dillwyn. however. 418 I'niri rsil y of California Publications in Botany [Vol.8 is convinced from the description of Dillenius, as well as by "the original drawing in Sir Joseph Bank's Library" that he is correct in his interpretation. Ectocarpus tomcntosus varies much in height and in the matter of the greater or less branching of the rope-like masses into which the erect filaments and their branches are intertwined. Our Pacific Coast plants are all shorter than the average and more simple. We have found, in certain specimens, long seriate intercalary gametangia of the type of those of Pylaiella near the tips of some of the branches. The occurrence of such gametangia in several species of true Ectocarpus is worthy of note as well as extremely puzzling. They were par- ticularly noticed also in specimens of E. confcrvoides f. variabilis. The name Conferva tomentosa first appears in 1762 in the earliest edition of Hudson's Flora Anglica (p. 480). Hudson quotes two numbers of Dillenius 's Historia Muscorum (1741), viz., no. 12 and no. 13 {Joe. cit., p. 19 and pi. 3, figs. 12 and 13). In the later editions of his work (1778, etc.) Hudson quotes only the second (no. 13) of these Dillenian descriptions as truly C. tomentosa., referring the first (no. 12) to his Conferva albida, now usually recognized to be Clado- phora albida Kuetzing. The actual founding of Conferva tomentosa in the sense of Dillwyn, therefore, rests on the synonymy quoted in the edition of 1778 which is, so far as we are aware, the only Hudsonian reference thus far quoted for the species. 5. Ectocarpus corticulatus Saunders Fronds 0.2-3 cm. high, tufted or feathery, from a small compact mass of creeping filaments; main filaments 90-120/* broad, usually densely corticated, irregularly and frequently branched ; cells of main filament 65-90/* broad, usually shorter than long, doliif orm ; cells of branches and ramuli seldom as long as broad, terminal cells tapering, blunt ; chromatophores band-shaped, few in each cell, often with pyrenoids ; sporangia unknown ; gametangia narrowly to broadly ovoid, 30-40/* long, 12-18/* broad, secund on the branches and ramuli, or even on the corticating filaments, short-stalked or sessile. On larger Melanophyceae and on Zostera. Alaska (Popof Island) to southern California (San Pedro). Saunders, Phyc. Mem., 1898, p. 152, pi. 20. Ectocarpus confcr- voides corticulatus Saunders, Alg. Harriman Exp., 1901, p. 418; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 238; Collins, Mar. 1! ' 23 ] Setchell^Gardner: Melanophyceae 419 Alg. Vancouver Is., 1913, p. 106, not Ectocarpus gramdatus I', corii- cuLatus Collins, in Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1590. There are two species of Ectocarpus of fair height with broader but short cells in the filaments occurring on our coast. One of these is E. acid us S. and G. and the other E. corticulatus Saunders. The differences between these tAvo usually strongly corticated species are to be found in the differing terminal cells of the branches and ramuli and in the shapes and dimensions of the gametangia. After carefully considering the different points of view as represented in the syn- onymy, we have felt compelled to recognize the autonomy of this specii-- as well as its close relationship with Ectocarpus arutus. 6. Ectocarpus fructuosus S. and G. Fronds tufted, profusely and alternately branched, up to 2.5 cm. high, attached by a mass of relatively short, creeping filaments; erect filaments several times forked near the base, producing long branches beset throughout with numerous, short, blunt ramuli of 2-3 orders; cells of main erect filaments cylindrical to slightly doliiform, and in part constricted at the cross-walls, 18-25^ diam. below, 2-4 times as long, slightly wider and shorter above; zoosporangia unknown; gametangia numerous, broadly conical, sessile or on 1-3-celled pedicels, 50-70/A long, 25-35/x broad at the base. Growing on the pneumatocyst of Nereocystis Luetkeana. Moss Beach, San Mateo County, California. Setchell and Gardner, Phyc. Cont, VI, 1922, p. 410, pi. 45, figs. 1-4. Ectocarpus fructuosus comes within the E. confervoides group, but seems amply distinct from any known forms to constitute a species. This species is based upon the character of the branches, viz., numerous long branches, producing throughout a great abundance of short, rather blunt ramuli and upon these very numerous, rather short and blunt, predominatingly sessile, gametangia. It seems to be a rare species. 7. Ectocarpus mucronatus Saunders Fronds loosely intertwined, 1-6 cm. high, light olive green, attached by a few colorless creeping filaments; erect filaments profusely branched in the upper parts, piliferous; primary branches sub- dichotomous ; ramuli alternate, mostly short and pointed ; cells of 420 University of California Publications in Botany [Vol. 8 main filaments cylindrical, 30-40/* diam., 1-3 times as long as the diameter below, 0.3-0.6 as long above; ehromatophores numerous, small, discoid ; zoosporangia unknown ; gametangia numerous, broadly ovoid, short-acuminate, sessile, 50-100/* (up to 155/*) long, 20-30/* broad. Growing on other algae and on Zostera. Vancouver Island (Bay- ard's Point) to southern California (San Pedro). Saunders, Phyc. Mem., 1898, p. 152, pi. 19. Ectocarpus granulosus Tilden, Amer. Alg. (Exsicc), no. 359& (not 359a,) • Ectocarpus mucronatus seems to be a species of frequent occur- rence in southern California. We have referred a very similar plant from the Puget Sound region (no. 359& of Tilden) here also. It is to be distinguished from other larger feathery species of this genus particularly by its broadly ovoid, abruptly and shortly mucronate gametangia. 8. Ectocarpus Taoniae S. and G. Plate 51, fig. 15 Fronds diffuse, 0.5-1.5 mm. high, attached by profusely branched, superficial, creeping filaments ; erect filaments simple, narrowed slightly at the base, gradually attenuated above, pilif erous ; cells of the creep- ing filaments cylindrical, 4-6/x diam., those of the erect filaments 8-10/* diam., quadrate below, 4-6 times as long as the diameter at the apices ; ehromatophores short, relatively thick, irregular bands; zoosporangia unknown; gametangia narrowly ellipsoidal, at times slightly curved, mostly sessile on the creeping and on the erect filaments, 20-28/* (up to 40/*) long, 15-20/* broad. Growing on Taonia Lcnneb acker ae. San Pedro, California. Setchell and Gardner, Phyc. Cont, VI, 1922, p. 413, pi. 46, fig. 15. This species forms rather diffuse and even layers, mostly along the margins of Taonia, over whose surface the creeping filaments spread and give rise to gametangia or to erect filaments not over 1.5 mm. high. In many ways it seems like a miniature Ectocarpus Padinae (Buffh.) Sauvageau (1896c-1897tt, p. 30 et seq.). Our species, how- ever, is only about one-half as tall, the cells of the erect filaments about one-half as broad, and the gametangia less than half as large as those of E. Padinae. We have found gametangia only of the "antheridium" type. Ectocarpus Padinae has its creeping filaments entirely endo- phytic while E. Taoniae has them entirely epiphytic. 1925 J Setchell-Gardner: Melanophyceae 421 9. Ectocarpus terminalis Kurtz. Fronds 0.75-2 (or 4) nun. high, mostly forming a continuous velvety layer, brownish in color; creeping filaments irregular, branched, anastomosing and forming, at times, a partially pseudo- parenchymatous layer ; erect filaments simple or sparingly branched, more or less attenuated above; cells of creeping filaments 8-24//. long, lO-lfy/, broad ; cells of erect filaments cylindrical, 8-12//, broad, up to six times as long; chromatophores short, band-shaped, few in each cell; zoosporangia terminal, ellipsoidal, 26-52//. long, 20-30//. broad; gametangia ovoid or ovoid-oblong, often curved, terminal, or lateral and sessile or short-stalked, 48-120/* long, 16-32//. broad. On larger Melanophyceae. Alaska (Unalaska) to southern Cali- fornia (Laguna). Kuetzing, Phyc. Germ., 1845, p. 236, Tab. Phyc, vol. 5, 1855, pi. 74, fig. Ill; Kjellman, Skand. Ectocarp., 1872, p. 54, pi. 2, figs, la, lb; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 237 ; Collins, Holden and .Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1034 (in part) (not no. 1387). We have accepted Kjellman 's interpretation of Kuetzing 's Ecto- carpus terminalis and refer here, although not without some hesita- tion, no. 1034 of the Phycotheca Boreali-Americana found growing on stipes of Maria fistulosa in Unalaska Bay. There is intermingled in our specimens, however, a plant which seems to be a form of Ectocarpus confervoides. We have also referred here a plant found growing on Fv.cus at Fort Ross, California. The creeping filaments in this species anastomose more or less and in some parts of the specimens form almost a parenchymatous basal layer, strongly resembling that of the species of the Myrionemataceae. Nearly every cell of the prostrate filaments gives off an erect filament, a condition characteristic among the Myrionemataceae, but not gen- eral among the Ectocarpaceae. The erect filaments are slender, with rather long cylindrical cells, and bear terminal zoosporangia and either terminal or lateral, sessile or short-stalked, gametangia. The chroma- tophores are clearly band-shaped though short. No. 1387 of the Phycotheca Boreali-Americana seems to be entirely made up of a form of Ectocarpus confervoides. 422 University of California Publications in Botany [Vol.8 10. Ectocarpus simulans S. and G. Fronds 1-2 mm. high, tufted; prostrate filaments irregular, tor- tuous, distinct ; erect filaments simple, tapering slightly upward, not piliferous ; cells cylindrical, not constricted at the dissepiments, ll-lSju diam. in the central part of the filament, 1-2.5 times as long as the diameter; chromatophores band-shaped, nearly covering the cell, few; zoosporangia unknown ; gametangia lateral or occasionally terminal, sessile, narrowly ellipsoidal, blunt, mostly slightly curved, 55-65/i long, 15-20/* wide. Growing on Chaetomorpha aerea. Cypress Point, Monterey County, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 412, pi. 45, figs. 9-11. The present species resembles, in general, certain forms of Ecto- carpus pusillus Griffiths (cf. Sauvageau, 1895), but the erect filaments are much more slender, usually never branched, and with much smaller gametangia. It forms tufts on Chaetomorpha aerea., whose filaments are encircled by the prostrate filaments of the Ectocarpus which form small cushions in their growth over one another. The gametangia are rather blunt and often more or less curved. 11. Ectocarpus flagelliferus S. and G. Fronds densely tufted, 3-5 mm. high, attached by a few creeping filaments and by small, colorless, descending rhizoidal filaments from the lower cells ; erect filaments simple below, sparingly and alternately branched above, tapering gradually upward and abruptly at the base, piliferous when young; cells of the main filaments slightly doliiform, 25-35//, diam., 0.25-1.5 times as long as the diameter ; chromatophores small, irregular bands or plates, without pyrenoids; zoosporangia unknoAvn; gametangia very variable in shape and size, fusiform to cylindrico-conical, mostly lateral on long pedicels, occasionally ter- minal on short, erect filaments, or long seriate-intercalary on the main filaments, the lateral 125-200/* long, 28-40/x, broad, the terminal and intercalary up to 1.5 mm. long. Growing on eel grass. Sitka, Alaska. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 408, pi. 47, figs. 25-27. 1925] Setchellr-Gardner: Melanophyceae 423 Ectocarpus flagelliferus , while resembling' superficially some of 1 1n- shortest plants of Ectocarpus confervoides f. variabilis, presents sev- eral peculiarities which are characteristic. The plants of this species are attached, so far as the adult specimens we have for study are concerned, by rhizoidal filaments which originate from several of the lower cells of the erect filaments and form complex basal masses which almost completely obscure the creeping filaments. Neither the rhizoidal filaments nor the creeping filaments, however, penetrate the host. The chromatophores are small and irregular in outline, seeming to be short bands rather than regular disks. They are generally so closely placed in the younger cells as to seem almost like a continuous band, but are separate in the older cells, with slender processes almost connecting them to one another. The cells are short and the erect filaments and their branches extend out into long hairs like whip- lashes. The lateral gametangia are variable in shape, arranged much as in Ectocarpus confervoides f. variabilis, but are of different dimen- sions. Besides the characteristic Ectocar pus-type of gametangia which are lateral, there occur terminal seriate gametangia of the Pylaiella-type which reach an extreme length of 1.5 mm. 12. Ectocarpus Mesogloiae S. and G. Fronds minute, 0.75-1.5 mm. high, attached by a mass of densely intertwined, branched, rhizoidal filaments penetrating among the cells of the host ; erect filaments sparingly and alternately branched ; main filaments and ramuli tapering gradually upward, very acute, not piliferous; cells cylindrical, slightly constricted at the dissepiments. 15-18/* diam. at the base, 1-2 times as long, 4-6/* at the apices of the filaments; chromatophores thin, irregular bands, nearly covering the cell; zoosporangia unknown; gametangia narrowly cylindrico-conical, 120-160/* (up to 210/*) long, 18-22/* broad, on short pedicels, rarely sessile. Growing on Myriagloia Andersonii (Farlow) Kuck. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 411, pi. 45. figs. 5, 6. The general characters of this diminutive species ally it with the E. confervoides group. We have deemed it best to consider it a dis- tinct species on account of the small dimensions of all of its parts and its penetrating habits. It has not been seen on any other host except the one mentioned above. 424 University of California Publications in Botany [Vol. 8 13. Ectocarpus commensalis S. and G. Fronds 200-400/*. high, densely fasciculate, attached by inter- twined, slightly branched rhizoids penetrating among the utricles and even well among the medullary filaments of the host ; erect filaments alternately branched near the base, simple above, slightly attenuated upward, not piliferous; cells of the erect filaments cylindrical, not constricted, 12-18/* diameter at the base of the filaments, 1.5-2.5 times as long as the diameter ; chromatophores band-shaped, few in each cell ; zoosporangia unknown ; gametangia blunt-fusiform to cylindrico- conical, short-pedicellate, mostly lateral near the base of the filaments, rarely terminal, 60-100/* long, 15-20/* wide. Growing on Codium Setchellii Gardner, Carmel Bay, and on C odium fragile (Suring.) Hariot, Pacific Grove, Monterey County, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 407, pi. 48, figs. 32-35. This minute species of Ectocarpus is one of several occurring on species of Codium, both on this coast and on that of southwestern Europe. The colorless rhizoidal filaments descend into the substance of the host even to the central (or basal) medullary tissues. The smaller diameter of the erect filaments and the more narrowly conical or fusiform gametangia readily distinguish this species from the others. 14. Ectocarpus eramosus S. and G. Fronds diffused or in diminutive tufts, 1-3 mm. high, attached by irregularly branched rhizoidal filaments penetrating deeply among the utricles of the host; erect filaments eramose or rarely producing short, divergent branches usually terminated in gametangia, slightly atten- uated toward the base and the apex, not piliferous; cells of mature filaments 28-40/* diam., 0.5-1.25 times as long as the diameter, cylin- drical, very slightly constricted; chromatophores relatively thick, closely crowded, irregular bands without pyrenoids; zoosporangia unknown; gametangia very variable in shape and size, lateral and alternate along the entire filament or occasionally terminal arid seriate on the main filaments, mostly on few-celled pedicels, lateral forms cylindrical to cylindrico-conical, 150-230/* long, 26-36/* diam., ter- minal, seriate forms, 700-900/* long, 15-20/* broad. 1925] Setchell-Gardner: Melanophyceae 425 Growing on C odium fragile in the lower littoral belt. Near the entrance to Tomales Bay, Marin County, California. SeteheU and Gardner. Phyc. Cont. VI, 1922, p. 407, pi. 47, figs. 18-23. This very small species has the general appearance of a form of Ectocarpus confer voides, but the cells are short and the filaments very slightly constricted at the partitions. It seems, in these respects, nearer to E. acutus and E. corticulatus, possibly bearing something of the same relation to these species that the dwarf forms of E. con- fervoides do to the typical form. Ectocarpus eramosus, however, is not readily to be referred as a dwarf form of either E. acutus or E. corticulatus and is consequently to be kept separate, at least for the present. 15. Ectocarpus luteolus Sauv. Prostrate filaments moniliform, irregularly branched, covering, or even occupying, the injured cells on the surface of the host, even occasionally penetrating slightly inward, not anastomosing, but form- ing an irregular and confused layer of almost parenchymatous aspect emitting tortuous rhizoidal filaments from its lower cells; erect fila- ments very short, 100-300//. high, terminating in delicate hairs (or, at times, in seriate gametangia), simple or with a few branches from near the base ; cells of the erect filaments short, swollen, and nearly globular below, cylindrical and longer above, about 8/t broad, 1.5 times as long as broad ; ehromatophores few in each cell, band-shaped ; zoo- sporangia terminal on short basal branchlets, oblong-ellipsoidal, 26-30/x long, 16-18/* broad ; gametangia narrowly cylindrical, con- sisting of single or double rows of loculi, 30-80/x (in ours about 30-45/*) long, 7-13/a (in ours 11-13//) broad. Forming small expansions on the surface of the lower portion of Pelvetiopsis limitata f. typica Gard. Central California (San Francisco). Sauvageau, Sur. quelq. alg., 1892, p. 79 (p. 42 in repr.), pi. II, figs. 14-19; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1233. Streblonema lutecium De-Toni, Syll. Alg., vol. 3, 1895, p. 575. Ectocarpus luteolus Sauv. is one of the species which seems inter- mediate between Ectocarpus and Streblonema. It does penetrate the host slightly and occasionally, but it seems to grow on the lower abraded portion of the host, filling the hollows between and even the I_(i University of California Publications in Botany [Vol.8 cavities of the injured cells. Its prostrate filaments and the lower portions of the erect filaments have moniliform, almost globular, cells which; together with rhizoidal filaments issuing from the very ba.sal ones, while perfectly separate, form a thin layer or cushion, with something of the appearance of that of Elachistea. Our specimens seem to agree so closely with the figures and descrip- tion of Sauvageau and to grow under such almost identical conditions, that we must necessarily refer them to this species. Thus far it has been detected only at Lands End, San Francisco, California, and only on one occasion. Our specimens show zoosporangia (not hitherto described) as well as gametangia. 16. Ectocarpus granulosus (Eng. Bot.) Ag. Fronds tufted, rather coarse and rigid, 1-8 cm. (up to 16 cm.) high, profusely branched ; primary branches opposite or whorled, mostly corticated below ; secondary branches opposite or at times alternate, short, tapering upward, often recurved at the tip, ending in hairs ; ultimate ramuli secund, short, acute, ending in short hairs ; cells of the main filaments 80-100/x diam., quadrate or shorter than the diameter, constricted at the cross-walls; chromatophores small, regular disks; zoosporangia "sessile, globose"; gametangia sessile, broadly ovoid, asymmetrical, mostly secund on the terminal ami subterminal ramuli, 60-100/x long, 30-6G> broad. Growing on various other algae and on stones in the lower littoral belt. Washington (Puget Sound) to southern California (San Pedro). Agardh, Sp. Alg., vol. 2, 1828, p. 45; Saunders, Phyc. Mem. 1898, p. 156, pi. 24, figs. 3-5 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 238; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1589. Ectocarpus granulosus f. corticulatus Collins, in Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1590. Conferva granulosa English Botany, 1814, pi. 2351. Ectocarpus granulosus was discovered by W. Borrer in the vicinity of Brighthelmston (Brighton), East Sussex, England, and is now reported from many coasts of Europe and both eastern and western North America. There is considerable variation, especially as regards dimensions both of the primary branches and of the gametangia. In some of our specimens, the gametangia reach a length of 165^, and a breadth of 65^, but the majority are 90-1 10ft long and 50-60/* broad. We have not seen zoosporangia in any of our specimens. The very 1925] Setchell-Gardner: Melanophyceae 427 distinctly discoid chromatophores and the opposite arrangemenl of most of the main branches clearly distinguish tliis species from any other on our coast. 17. Ectocarpus Parksii S. and G. Plate 49, fig. 15 Fronds 5-7 cm. high, densely caespitose, flaccid ; main filaments of erect fronds, 34-40/t diam. in the lower part, tapering very gradually upward, profusely and alternately branched, very slightly corticated at the base ; branches of succeeding orders reduced in diameter, long and tapering very gradually upward, not terminating in hairs; ultimate ramuli 10-15/* diam., terminal cell 5-7/* diam. ; cells cylin- drical to very slightly doliiform, 40-70/* long in the main filaments, quadrate to subquadrate above; chromatophores numerous, irregular plates densely crowded together and more or less connected in the ramuli, becoming more regularly disk-shaped and separate in the lower parts of the main filaments ; gametangia relatively sparse, sessile, straight or slightly curved upward, mostly on the subterminal and terminal ramuli, secund, 4-10 in a group, occasionally solitary, blunt fusiform to subcorneal, 35-45//. long, 18-22/* broad. Growing on Gracilaria confervoides (L.) Grev. in a small cove, San Francisco Bay, northeast of Tiburon, Marin County, California. Setchell and Gardner, Phyc. Cont. VII, 1924, p. 1. This species of Ectocarpus appears to be a very short lived spring form. Within a month after it was first observed, apparently just coming into fruit, it had largely disappeared. Its structure seems to relate it to several of our west coast species. Its long, slender, grad- ually tapering filaments, profusely branched, are similar to those of E. silicxdosxis (Dillw.) Lyngb. Its sessile gametangia are similar in form to those of E. Taoniae S. and G. The secund arrangement of the gametangia and the disk-shaped chromatophores are characters belong- ing to E. granxdosoides S. and G. The delicate fronds, with profuse, alternate branching, the disk-shaped chromatophores, and the groups of small, sessile, blunt, secund gametangia make a combination of characters not found in any other species. 428 Uwi/versity of California Publications in Botany [Vol.8 18. Ectocarpus Mitchellae Ilarv. Fronds aggregated into dense feathery tufts, 2-8 cm. high, decom- poundly branched, attached by long creeping filaments; yellowish green or olive green on drying; branching alternate, the branches divaricate, attenuated, often ending in short hairs, ultimate ramuli approximated; cells in the main branches 25-40./*. diam., 1-3 times as long as the diameter below, shorter above ; chromatophores numerous, small, discoid ; zoosporangia unknown ; gametangia ellipsoidal to nar- rowly oblong, obtuse, 50-100/t long, 18-35/t wide, sessile, mostly secund on the upper side of the branches. Growing on limpets and on Phyllospadix in the lower littoral belt. Southern California. Harvey, Ner. Bor.-Amer., part I, 1851, p. 142, pi. 12, G; Collins, Notes on New England Algae V, 1891, p. 337 ; Saunders, Phyc. Mem., 1898, p. 153, pi. 21, figs. 1, 2; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 671. We have specimens from several collections from the coast of southern California which agree so well with plants from the southern New England coast referred to E. Mitchellae that we cannot satisfac- torily separate them from it. We call attention to the views of Sau- vageau (1896&, p. 39), Collins (1891, p. 337), Saunders (1898, p. 153) and Borgesen (1914, pp. 159-162) as to the possible relation, or even identity, of this species with E. indicus Sond., E. virescens Thuret, and E. Duchassaingianus Grunow, but we have no additional facts to present. Our specimens show both mega- and meio-gametangia such as Sauvageau (18966) describes and illustrates for E. virescens Thuret. 19. Ectocarpus oviger Harv. Fronds somewhat densely matted together, 7-15 cm. (up to 22 cm.) high, main filaments densely corticated, dark brown, decompositely branched ; branches mainly alternate, in part secund, strict ; cells cylindrical throughout the frond, not constricted, 60-70/x diam. in the main filaments, 1.5-2.5 times as long as the diameter; chromato- phores discoid, numerous ; zoosporangia uncertain ; gametangia broadly ovoid to subspherical, asymmetrical, sessile or subsessile, alter- nate or in part secund, promiscuously scattered throughout the branches, 35-45/1 long, 30-35/t broad. i 925 J Setchellr-Gardner: Melanophyceae 429 Growing on Nereocystis Lu< tkcana. Puget Sound region to central California. Harvey, Coll. Alg. N.W. Amer., 1862, p. 167 ; Setchell and Gardner. Alg. N.W. Amer., 1903, p. 238. Ectocarpus granulosus Tilden, Amer. Alg. (Exsicc.), no. 359a (not no. 3596) ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 2242. The Ectocarpus oviger Harvey has been a puzzle for phy eulogists who, in general, have been inclined to refer it to E. granulosus, to which latter species Harvey himself compared it. An examination of the type specimen at Dublin and a careful study of Harvey 's descrip- tion, lead us to refer here no. 359a of Tilden 's American Algae and no. 2242 of the Phycotheca Boreali-Americana and to keep E. oviger, at least as thus interpreted, distinct from E. granulosus. Harvey speaks of his plants as having the aspect of Pylaiella littoralis, and the plants we refer to his species resemble certain varieties or forms of that species more than they do any forms of E. granulosus. Harvey also remarks that the "fruit" of his species is like that of E. granu- losus, and it is true that, in our specimens, which show only game- tangia, these resemble the gametangia of E. granulosus more than those of any other species, but differ from them sufficiently in shape and dimensions to prevent merging the one species into the other. Our conclusion is that Ectocarpus oviger, while closely related to E. granulosus, is sufficiently different from it in habit, in the length of the main filaments, and in the shape and size of the gametangia, to be retained as an independent species. "We have indicated that we are uncertain as to the zoosporangia in this species, although it has been supposed that the ' ' fruit ' ' described by Harvey is zoosporangial. The bodies seen on the portion of the type specimen examined seemed to be empty zoosporangia. On the other hand, the gametangia at maturity and about to discharge their gametes show little or no trace of the internal partitions, so that further experience with this species is very desirable. 20. Ectocarpus breviarticulatus J. Ag. Fronds tufted, 2-4 cm. high, arising from creeping filaments ; erect filaments about 27/a broad (35-45/x, fide Kuetzing and De-Toni), either sparingly branched or more or less beset with short branchlets, both branches and branchlets attenuated toward the apex and hooked, intertwined into rope-like masses; cells usually 1.5-2 times as long as 4:!0 University of California Publications in Botany L VoL - 8 broad; chromatophores disk-shaped, several in each cell; zoosporangia unknown; gametangia from nearly spherical to broad ovoid, about 62^ long, 57/A broad. On Chnoospora pacified. West coast of Mexico (St. Augustin). J. G. Agardh, Nya Alger fran Mexico, 1847, p. 7, Sp. Alg., vol. 1, 1848, p. 16 ; Kuetzing, Spec. Alg., 1849, p. 453 ; Borgesen, Mar. Alg. Dan. West Indies, pt. 2, 1914, p. 173, fig. 136. Ectocarpus hamatus Crouan, in Maze et Schramm, Ess. Class. Alg. Guad. (ed. 2), 1870-77, p. Ill (fide Borgesen) ; Vickers, Phyc. Barbad., pt. 2, 1908, pi. 29. Ectoca/rpus brcviarticidatus J. Ag. is known to us as a member of our flora only from the literature and particularly from Borgesen 's (he. cit.) account of it. It seems closely related to E. oviger, from which it differs chiefly in size and branching, and slightly in dimen- sions of filaments and gametangia. Both Kuetzing and Borgesen state that they have examined cotypes and may be, therefore, con- sidered to have spoken with authority. We have been enabled to study what certainly seems to be this species in material from American Samoa (cf. Setchell, Veg. Tut. Isl., p. 171, fig. 37). 21. Ectocarpus chantransioides S. and G. Fronds arising from contorted creeping filaments, forming dense hemispherical cushions 4-8 mm. high ; branching profuse, alternate below, mostly secund above; main filaments and branches not atten- uated; cells 8-10/x diam., 2-3 times as long as the diameter below, quadrate above ; chromatophore band-shaped ; zoosporangia unknown ; gametangia mostly sessile or on short pedicels, narrowly cylindrico- conical, 80-110/* long, 16-20//. broad at the base. Growing on boulders in the lower littoral belt. Three miles north- west of Santa Monica, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 406, pi. 48, figs. 28-31. The habit of Ectocarpus chantransioides distinguishes it at once from all other species of the genus with the exception of E. hemi- sphericus Saunders. The latter species is always found, so far as our knowledge is concerned, epiphytic on Fucaceae, while the former species is confined to rocks. Ectocarpus chantransioides has also more slender filaments than has E. hemisphericus, not at all tapering, and has distinct apical growth. The gametangia are differently shaped, being longer and more slender. It is therefore very distinct even from E. hemisphericus. It resembles the genus Choristocarpus of the l925 ] Setchellr-Gardner: Melanophyceae 431 family Choristocarpaceae, as diagnosed by Kjellman (1891, pp. 190, 191), in having apical growth. The growth in length of the erect filaments in the Choristocarpaceae is by the division of the terminal cell only, whereas in E. chantransiuuh s the meristem extends over a number of cells at the outer or apical end of the filaments. These cells, some 10 to 15 in number, are much richer in cell contents, the terminal cell being the richest of all in the series. This is a xrvy unusual condition for an Ectocarpus. The nearly uniform diameter throughout of the erect filaments and their method of branching resemble to a remarkable degree those characters found in Chan- transia. The gametangia are typically those of Ectocarpus and, not- withstanding the other rather unusual characters, it seems to be most closely related to that genus, but a very distinct species. 22. Ectocarpus granulosoides S. and G. Fronds 2-3 cm. high, profusely branched; main filaments sub- dichotomously branched, all branches suddenly attenuated at the base, corticated below-, secondary filaments mostly alternate, in part secund, strict, long-attenuate, acute, not piliferous ; ultimate ramuli mostly secund, acute; cells of the main filaments 70-80/* diam., 0.5-1 times as long as the diameter, those of the secondary filaments 30-40ju diam., 0.3-1 times as long and those of the ultimate ramuli 14-20/x diam.. 0.3-1.5 times as long, all slightly constricted at the dissepiments ; chromatophores numerous, discoid in the' older cells, irregularly angular in the ramuli ; cells of the corticating filaments 7-10/t diam., 3-4 times as long as the diameter ; zoosporangia unknown ; gametangia secund on the upper side of the ultimate and penultimate ramuli, sessile, subfusiform, asymmetrical, 40-60/*. long, 12-20/* broad. Growing on rocks (?). San Pedro, California. Setchell and Gardner, Phyc. Cont, VI, 1922, p. 410, pi. 45, figs. 7, 8. The very distinct species we have described under the name of Ectocarpus granulosoides has the habit of a small E. granulosus and gametangia of the same general type as found in that species, but considerably smaller. The branching, however, is never opposite, and the acute ramuli gradually attenuated upward recall those of E. acutus, from which our species is amply distinct in both chromato- phore and characters of its gametangia. The branches, particularly the larger, are very suddenly and considerably attenuated at the base, giving the species a striking characteristic of its own, at least within the group of species with discoid chromatophores. 432 University of California Publications in Botany [Vol. 8 23. Ectocarpus affinis S. and G. Plate 51, figs. 16, 17 Fronds diminutive, 500-700/4 high, attached to the host by a few branched, colorless, entwining, rhizoidal filaments; erect filaments sparingly branched, gradually attenuated upward, piliferous; branches alternate or opposite ; cells of the main filaments cylindrical, 24-30**. diam. at the base, 9-11/*, at the apex, varying in length throughout the filament, 0.75-3 times as long as the diameter; chromatophores small, thin, angular plates in young cells, nearly regular disks in older cells ; zoosporangia unknown ; gametangia lateral, mostly sessile, occasionally on 1-celled pedicels, solitary, secund, or in whorls from a single cell. blunt-conical to spheroidal, 28-31/*. long, 25-28/* broad, arranged along the entire filament to the terminal hairs. Growing on CaUithamnion sp. Sitka, Alaska. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 405, pi. 46. figs. 16, 17. Two collections from Sitka, Alaska, but both made on the same day, show a small Ectocarpus growing on a species of CaUithamnion which is so closely related to E. ovattis Kjellman (1877a, p. 35) that perhaps it may seem necessary, at some time, to refer it to that species. The plants are, however, less stout than those of Kjellman and with more nearly spherical gametangia. For these reasons and because of its geographic remoteness and the somewhat different climatic con- ditions, we feel that it is desirable to describe the Alaskan plant as a distinct, but closely related, species. 24. Ectocarpus cylindricus Saunders Fronds diffuse, 1-2 mm. high ; creeping filaments superficial ; erect filaments very sparingly branched above, several arising from the same creeping filament, tapering slightly at the base and apex; cells of the creeping filaments 16-20/x diam.; cells of the erect filaments cylindrical, slightly constricted at the cross-walls, 2-3 times as long as broad below and at the apex, 0.5-1 times in the middle of the filament ; gametangia mostly on the middle and upper parts of the filaments, usually alternate. Growing on Egregia Menziesii and Cystoseira osmundacea. Central California (Pacific Grove). 1925 ] Setchell-Gardner: Melanophycecu 4:5:} Saunders, Phyc. Mem., 1898, p. 150, pi. 16; Setchell and Gardner, Phyc. Cont. VI, 1922, p. 415 | f. typicus). The typical form, growing on the larger brown algae, differs some- what from the variety found growing on C odium as is indicated below. Ectocarpus cylindricus f. codiophilus S. and G. Plate 51, fig. 14 Fronds densely eaespitose, 3-5 mm. high; rhizoidal filaments densely intertwined, penetrating the host; erect filaments anastomos- ing and forked at the surface of the host, unbranched above, pili- ferous; gametangia, for the most part, near the base of the erect fronds. Growing on Codium fragile (Suring.) Ilariot and ('odium Setchellii Gardner. Carmel Bay, Monterey County, and La Jolla, San Diego County, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 415, pi. 46, fig. 14 and pi. 49, figs. 42-45. This form differs from the typical form in having penetrating rhizoidal filaments and in having the gametangia in a dense zone at or near the base of the erect filaments instead of being scattered. Ectocarpus cylindricus f. acmaeophilus S. and G. Fronds tufted, 7-10 mm. high ; creeping filaments superficial ; erect filaments unbranched, nearly the same diameter throughout ; zoosporangia unknown ; gametangia up to 270/* long, mostly opposite on the upper part of the erect filaments. Growing on Acmaea sp. Carmel Bay, Monterey County, Cali- fornia. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 415, pi. 49, fig. 46. This form seems to be rare, at least but few specimens have been detected up to the present time. It differs from the typical form in being practically unbranched, except the fructiferous branches. Where branches occur, they seem to arise as the result of an injury to the main filament. Also it differs in having the erect filaments more nearly cylindrical throughout. And finally it differs in having the gametangia more often opposite, and generally longer and proportionally narrower. 434 University of California Publications in Botany [Vol.8 25. Ectocarpus hemisphericus Saunders Fronds pulvinate, 2-4 mm. high, arising from a compact network of creeping filaments rhizoidal in character, much attenuated and crisped ; primary branches pseudodichotomous, divaricate, gradually attenuated upward ; secondary branches numerous, short, clustered, mostly long-piliferous; cells at the base of main filaments 22-26//. diam., 1—3 times as long as the diameter, smaller above, slightly con- stricted at the joints ; chromatophores numerous, small, discoid ; zoo- sporangia cylindrical, 30-35/*. long, 20-25// diam., on 1-celled stalks, at times on the same fronds with the gametangia (Saunders) ; game- tangia fusiform to broadly ovoid, obtuse, 30-90//. long, 14-20// diam., lateral on the upper parts of the fronds, short-pedicellate. Growing mostly on Pelvetia fastigiata, but also on other species of algae. California (San Diego, Laguna and San Pedro). Saunders, Phyc. Mem., 1898, p. 151, pi. 17 ; Guernsey, Notes on Mar. Alg., 1912, p. 198, fig. 108 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 528. Ectocarpus hemisphericus f. minor Saunders Very similar to the species, but forming less definite tufts and with the erect filaments simple or but slightly branched, 1-2 mm. high, 16-21// diam., and with gametangia and zoosporangia mostly sessile. Growing on and near the receptacles of Hesperophycus Har- veyanus. San Pedro, California. Saunders, Phyc. Mem., 1898, p. 151, pi. 18, figs. 1-3; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), nos. 529 and 1794. Ectocarpus hemisphericus Saunders, in typical form, can readily be recognized by its habit and its host, but in some forms and par- ticularly in f. minor is less distinct in habit, at least. Thus far, both the species and the form seem confined to the coast of southern California. 26. Ectocarpus Saundersii S. and G. Fronds forming small tufts, 2-5 mm. high, attached by a mass of branched filaments creeping on the surface of the host ; branching alternate ; main branches long and divergent, ultimate ramuli sparse, short, acuminate; cells of the main filaments 25-40// diam., quadrate below, shorter above, slightly constricted at the dissepiments ; chroma- 1925] Setchell-Gardner: Melanophyceae 435 tophores numerous, discoid ; zoosporangia on short stalks or inter- calary, globose, about 30/i diam. ; gametangia cylindrical to ovoid, obtuse or acuminate, lateral, borne on longer or shorter pedicels, 70-150^i long, 2d-50/jl broad. Growing on Fucus, in the middle of the littoral belt. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 411. Ectocarpus paradoxus var. pacificus Saunders, Phyc. Mem., 1898, p. 152, pi. 18, figs. 4-7; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 530. According to Saunders, his plant, the var. pacificus, differs from the type of E. paradoxus Mont., in lacking any manifestation of opposite branching as well as in having the gametangia longer and more pointed. In these respects and also because it is a shorter plant, it seems to us to be sufficiently distinct to be separated specifically from the type, and to be different from any other species known to us. 27. Ectocarpus acuminatus Saunders Fronds diminutive, attached by a network of delicate rhizoidal filaments ; erect filaments uniform in size throughout, 1-2 mm. long ; cells cylindrical, not constricted, 12-14/x diam., quadrate below, 2-5 times as long above; chromatophores irregular or elliptical disks, more abundant in the central part of the filament than elsewhere; zoo- sporangia unknown; gametangia sessile on the creeping filaments or on the base of the erect filaments or terminal on short filaments arising directly from the creeping filaments, very long-acuminate, sometimes tipped with a short hair, often more or less curved, 90-300^ long, 20-30/* broad. Growing within the coneeptacles of Cystoseira osmundacea and Halidrys dioica. Central California (Pacific Grove) to southern California (San Pedro). Saunders, Phyc. Mem., 1898, p. 149, pi. 14, figs. 1-5; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), nos. 524 and 2142. The shape of the gametangia would seem to ally this very incon- spicuous species most closely with E. siliculosus, but the chromato- phores are described by Saunders as elliptical. We have only dried material for examination and cannot, therefore, add anything to the description by Saunders. It is a very curious and seemingly distinct species. Its habitat, within the coneeptacles of members of the Fucaceae, seems to have reduced it almost to the state of parasitism, at least one-half of the plant, being unable to function in food making. 4:!(i University of California Publication* in Botany [Vol. 8 28. Ectocarpus ellipticus Saunders Fronds minute, about 1 mm. high, tufted, attached by numerous, colorless, branched rhizoidal filaments ; erect filaments sparingly dichotomous at the base, unbranched or with few unicellular branches above, gradually attenuated upward, piliferous ; cells 9-12/x diam. and 1-2 times as long as the diameter at the base of the filaments, 3 times as long above; "chromatophores thin branched bands, oval," numerous ; zoosporangia sessile, ovate or ellipsoidal, 30-40/a long, 12-18/x broad; gametangia cylindrical to blunt-conical, 75-100^, (up to 180/x.) long, 18-25/1. diam., borne laterally at or near the base of the erect filaments, or occasionally arising directly from the creeping filaments, sessile or on short pedicels below ; loculi one to few seriate. Growing in the conceptacles and cryptostomata of Fucus sp. Pacific Grove, California. Saunders, Phyc. Mem., 1898, p. 149, pi. 14, figs. 6-9; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 527. Ectocarpus ellipticus is a curiously localized species growing upon the inner margins of the openings of the cryptostomata and con- ceptacles of Fucus at Pacific Grove, California. Transverse sections of the fronds of the Fucus show that this Ectocarpus simply forms a ring of growth just within the outer opening of the cryptostomata and conceptacles just outside the hair-forming tissues. The rhizoidal fila- ments of the Ectocarpus penetrate into the apparently looser tissues of the ostiole of the cavities it inhabits. The gametangia. are all basal or borne near the bases of the erect filaments. It seems to be a very distinct species. 29. Ectocarpus chitonicola Saunders Fronds small, 1-2 mm. high, tufted, creeping filaments numerous, irregular, branched ; cells of the creeping filaments 11-15/x diam., 1-2 times as long as the diameter ; erect filaments mostly simple, 14/x diam. at the base, somewhat narrower above, with cells 0.5-2 times as long as the diameter below, 2-3 times as long above, not constricted at the partitions ; chromatophores small oval disks, numerous ; zoosporangia unknown ; gametangia subcylindrical to narrowly ovoid, obtuse at the apices, 90-175/i. (up to 250/a) long, 25-35/a diam., borne laterally on the erect filaments, or occasionally on the creeping filaments, the lower gametangia on pedicels, the upper sessile. 1925] Setchell-Gardner: Melanophyceae 437 Growing: in minute tufts on the shells of Chitons and Limpets. Pacific Grove and Carmel Bay, California. " Ectocarpus chitonicolus" Saunders, Phyc. Mem., 1898, p. 150, pi. 15, figs. 1-4. The illustrations of Saunders (he. eit.) show a diversity of forms of the gametangia, ranging from long ones with many small loculi, to short ones with few large loculi. The species either possesses great range of variation in this respect, or else there may possibly be two species concerned. Another explanation of the phenomenon may be the possibility that we have in this species the two forms of game- tangia mentioned by Sauvageau (1896b, p. 17) found on Ectocarpus rirescens. Saunders' figure 2 might well represent the "mega- sporangia'' and figures 3 and 4 the "meiosporangia" of Sauvageau. The species needs further investigation to have the validity of these points established. In the specimens we have examined, both that of Saunders and the single one of Gardner's collecting, the gametangia are long, and large-celled. The species is to be distinguished from any other of our low forms by its discoid chromatophores and its peculiar habitat. 30. Ectocarpus socialis S. and G. Plate 51, figs. 12, 13 Fronds caespitose, 1-3 mm. high, attached by densely intertwined, penetrating rhizoidal filaments; erect filaments several times forked at the surface of the host, simple above or rarely alternately branched, long-attenuate upwards, piliferous; cells cylindrical to slightly doli- iform, not constricted, 22-28/* diam., 1-2 times as long as the diameter below, 4-6 times above ; chromatophores small, few in a cell, discoid ; zoosporangia ellipsoidal, erect, lateral, alternate, mostly on 1-2-celled pedicels, occasionally sessile near the base of the erect fronds, 60-95/i. long, 35-50/* broad; gametangia fusiform, sparse on the upper parts of the erect filaments; sessile or on 1-2-celled pedicels, alternate, 70-85/t long, 25-30/* broad. Growing on C odium fragile (Suring.) Hariot, in company with Ectocarpus globifer Kuetz. and with various species of Myxophyceae and Rhodophyceae. Redondo, California. Setchell and Gardner, Phyc. Cont. VI, 1922, p. 412, pi. 46, figs. 12, 13. Eetoearpus soeiaJis seems to be a relative of the E. cylmdricus group. The chief distinctions between the two species lie in the shape, size and position of the gametangia. 438 University of California Publications in Botany [Vol.8 31. Ectocarpus flocculiformis S. and G. Fronds densely caespitose, 0.75-1.5 mm. high, attached by inter- twined, penetrating, rhizoidal filaments; erect filaments more or less forked at the base, simple above or the largest plants with few, short, alternate, acute ramuli, tapering slightly at the base, long-attenuate upwards, not piliferous; cells cylindrical to slight doliiform, 20-25/* diam., 1-2.5 times as long ; chromatophores numerous, angular-discoid in the younger cells, rounded in the older cells; zoosporangia ellip- soidal, 56-66/* long, 35-4% broad, on short pedicels near the base of the erect filaments ; gametangia ovoid to ellipsoidal, numerous, sessile or short-pedicellate, lateral near the base of the erect filaments, 95-110/* (up to 150/0 long, 30-45/*. broad. Growing on Codium fragile, the rhizoidal filaments penetrating deeply into the host, among and beyond the utricles. La Jolla, California. Setchell and Gardner, Phyc. Cont, VI, 1922, p. 409, pi. 47, fig. 24. The nearest relative of E. flocculiformis apparently is E. cylin- dricus f. codiophilus. One marked difference between it and any of the forms of E. cylindricus is in the shape and size of the gametangia. This difference, along with others of less importance perhaps, though apparently constant, seems to be sufficient to warrant keeping it separate. 32. Ectocarpus globifer Kuetz. Fronds 2-5 mm. high, pulvinate, sparingly branched, attached by a few branched, rhizoidal filaments penetrating among the cells of the host ; branches opposite or alternate, divaricate, constricted at the base, attenuated above, at times piliferous; cells cylindrical, slightly constricted at the dissepiments, 45-50/* diam. in the main filaments, 0.5-1.5 times as long as the diameter in the middle, up to 4 times the diameter above and below; chromatophores numerous, small, irregular discoid; zoosporangia lateral, spherical or subspher- ical, mostly on 1-celled pedicels, 30-40/* diam.; gametangia lateral, broadly ovoid to spherical, mostly on 1-celled pedicels, 70-100/* long, 50-70/* broad. Growing on Codium fragile. Southern California (San Pedro and La Jolla). 1925] Setchell-Gardner: Mclanophyceae 4:5!) Kuetzing, Phye. Gen., 1843, p. 289, Tab. Phyc, 1855, vol. 5, p. 16, pi. 49, fig. 2; Bornet, Note sur quelq. Ect., 1891, vol. 38, p. 358 (repr. p. 6), figs. 6, 7; Collins, Holden and Setchell, Phye. Bor.-Amer. (Exsicc.),no. 732. This species has been distributed from southern California under no. 732 of the Phycotheca Boreali-Americana. The specimens dis- tributed were collected at La Jolla by Mrs. Snyder. Gardner has collected it twice at Redondo. Both undoubted zoosporangia and gametangia have been found. All of the specimens from our coast, so far as is known, grow upon Codium fragile, into whose spongy sub- stance they penetrate by long colorless rhizoidal filaments. The specimens agree so well with the figures and descriptions of Kuetzing and Bornet that there seems to be little doubt that our plants are of the same species as those from Europe. 33. Ectocarpus gonodioides S. and G. Fronds minute, forming small tufts 500-550/x high, attached by long, more or less hyaline rhizoidal filaments penetrating the host; filaments sparsely branched at the surface of the host, tapering rather abruptly at the base, long attenuated upward to a blunt apex, 18-24/t diam. at the base, 10-14/x at the apex ; cells 1-2 times as long as broad ; zoosporangia unknown; gametangia narrowly fusiform on 1-2-celled pedicels, near the base of the erect filaments, up to 125/* long, 20-28/* diam. in the widest part. Growing on Codium cuniatum. Smith Island, Gulf of California. Setchell and Gardner, The Mar. Alg. Gulf of Calif., 1924. pp. 721, 722, pi. 17, fig. 44. The small tufts which this species of Ectocarpus produces remind one of the genus Gonodia (Myrimtis), but the penetrating part, which extends relatively deep into the host, is composed of slender, almost colorless, slightly branched, closely intertwined filaments, which, how- ever, do not coalesce or form a false parenchyma as in the case of some species of Gonodia, The plants, though sparse, are in excellent fruit- ing condition. The chromatophores are too much disorganized for characterization. 440 University of California Publications in Bala an [Vol.8 34. Ectocarpus Bryantii 8. and G. Fronds intertwined, forming a more or less continuous stratum, 1-2..") mm. high, attached by relatively short, penetrating, rhizoidal filaments; erect filaments forked more or less at the surface of the host, with very few short ramuli above, nearly cylindrical, tapering slightly above, uncorticated ; terminal cell blunt, 28-32//. diam., cells 1-2 times as long as broad ; chromatophores small disks ; zoosporangia unknown ; gametangia narrowly to broadly fusiform, sessile or on 1-celled pedicels, 70-100// (up to 140//) long, 25-35//. broad, scattered promiscuously along the whole length of the erect fronds. Growing on C odium Brandegeei. La Paz, Lower California. Setchell and Gardner, Mar. Alg. Gulf of Calif., 1924, p. 720, pi. 17, fig. 45. Ectocarpus Bryantii and E. gonodioides are evidently closely related to each other and both have near affinities in the pusillus group of Sauvageau (1895). They both differ from all of the forms pro- posed in the method of branching and in having no hairs terminating the erect filaments. 5. Streblonema Derb. and Sol. Fronds composed of more or less branched, monosiphonous or in part polysiphonous filaments, wholly or largely endophytic ; prostrate primary filaments wholly endophytic, creeping among the cells of the host, erect secondary filaments wholly or in large part endophytic, simple or branched, hairs present or absent ; zoosporangia and game- tangia both present, terminal or lateral on the erect or on the prostrate filaments, sessile or, more rarely, short-stalked. Derbes and Solier, in Castagne, Supplem. Catal. Marseille, 1851, p. 100. The type species is S. sphaericnm. Key to the Species 1. Fronds causing noticeable distortions of the host 2 1. Fronds not causing noticeable distortions 3 2. Distortions in the form of pustules 13. S. scabiosum (p. 450) 2. Distortions in the form of extended rugose areas ...12. S. rugosum (p. 449) 3. Fronds producing noticeable patches or discolorations 4 3. Fronds inconspicuous 11 4. Patches large or extended 5 4. Patches small, usually orbicular (0.5 cm. or less diam.) 6 5. Patches orbicular 11. S. evagatum (p. 449) 1925] Setrhrll-Cimhii r: M elanophyceae 441 5. Patches indefinite in outline and extent 6. S. penetrale (p. 446 6. Patches definitely eruinpent 7 G. Patches discolored, filaments only slightly erumpent 7. S. transfixum (p. I Ml 7. Gametangia pluriseriate 8. S. irregulare (p. 447) 7. Gametangia uniseriate 8 8. Parasitic on Zostera 4. S. vorax (p. 444) 8. Parasitic on other algae 9 9. Spots aecidioid 14. S. aecidioides f. pacificum (p. 450; 9. Spots not aecidioid 10 10. Gametangia cylindrical 15. S. myrionematoides (p. 452) 10. Gametangia narrowly elliptical 10. S. pacificum (p. 448) 11. Erect filaments not over 8/u in diameter 12 11. Erect filaments over 10^ in diameter 15 12. Gametangia single 5. S. Porphyrae (p. 445) 12. Gametangia corymbose 13 13. Gametangia 25-40 M long 14 13. Gametangia over 60^ long 9. S. minutissimum (p. 447) 14. Erect filaments 4-5m in diameter 1. S. corymbiferum (p. 441 I 14. Erect filaments 5-8m in diameter 16. S. investiens (p. 452) 15. Gametangia 50-100m long, 10-16/* broad 2. S. anomalum fp 442) 15. Gametangia 90-130 M long, 28-36^ broad 3. S. Johnstonae (p. 444) 1. Streblonema corymbiferum 8. and G. Plate 52, fig. 8 Fronds microscopic, composed of irregularly and alternately branched filaments penetrating among the cells of the host ; fructi- ferous ramuli decidedly aggregated, forming corymb-like clusters near the surface of the host ; cells mostly cylindrical, in part irregular, 4-5/x diam., 1.5-4 times as long; chromatophore band-shaped, not covering the cell ; zoosporangia unknown ; gametangia cylindrical to slightly fusiform, blunt, terminal on short pedicels, borne in clusters, or sessile on filaments creeping near the surface of the host, 25-35/* long, 4.5-5.5/* broad; loculi uniseriate, dividing walls frequently oblique. Growing on Cumagloha Andersonii (Farlow) S. and G., in com- pany with Streblonema anomalum and 8. Johnstonae. San Pedro, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 391, pi. 43, fig. 8. While studying material of Cumagloki Andersonii (Farlow) S. and G. collected by Mrs. H. D. Johnston at San Pedro, California, in 1899 and deposited in the Herbarium of the University of Cali- fornia under no. 94663, some specimens of Ectocarpaceae were inci- dentally observed-. When we came to study our material of this 442 University of California Publications in Botany [Vol.8 family, these specimens were given a careful scrutiny with the result that three species new to science were detected on a single specimen of the host, viz., Streblonema corymbiferum, S. anomalum, and 8. Johnstonae, all of which have been published (loc. tit.). The habit of each and the size of all the parts are so characteristic that, notwithstanding their intimate association on the same host, they can readily be distinguished. S. corymbiferum is the most delicate of the three. The very frequent grouping cf the gametangia into corymb-like clusters is one of the most prominent distinguishing characters of the species. Other species of this genus are as small or even smaller. In size of parts, somewhat in habit and in habitat, it resembles S. mimdissimum Saunders found growing at Sitka, Alaska, on "Liebmannia sp." (Saunders, 1901, p. 416.) 2. Streblonema anomalum S. and G. Plate 52, figs. 1-3 Fronds microscopic, more or less confluent ; filaments of the prostrate portion penetrating among the cells of the host, moderately and irregularly branched, much contorted, up to 16/x diam. ; erect filaments unbranched, mostly cylindrical, blunt in part, and piliferOus in part, 200-250/*. long, extending beyond the surface of the host ; hair filaments sparse, arising on the creeping filaments; cells of the creeping filaments more or less divided 1-2 times lengthwise in the oldest parts, producing a polysiphonous filament 18-24/* diam. ; other cells somewhat irregular in form, 8-11/* diam., at times slightly con- stricted at the cross-walls ; cells of the erect filaments cylindrical, 8-9/*. diam., quadrate ; chromatophore a single thin parietal band, unbroken in the erect filaments, more or less broken in the creeping filaments ; zoosporangia ellipsoidal, ovoid or nearly spherical, 40-60/* long, 28-35/* broad, sessile on the creeping filaments ; gametangia cylindrico-conical, mostly blunt, 50-100/* long, 10-16/* broad, sessile or short-pedicellate on the creeping filaments, single or secund. Growing in Cumagloia Andersonii (Farlow) S. and G. in com- pany with Streblonema corymbiferum and S. Johnstonae. San Pedro, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 392, pi. 43. figs. 1-3. Streblonema fasciculatum Saunders, Phyc. Mem., 1898, p. 148, pi. 13 (not of Thuret, in Le Jolis, Alg. mar. Cherb., no. 100, Liste, 1863, p. 73). 1925] Setchell-Gardner: Melanophyceae 443 Examination of a small piece of authentic material of 8. fascicu- latum Saunders collected at San Pedro, California, in August, 1896, revealed the presence of specimens of the same three species mentioned under 8. corymbiferum, collected by Mrs. Johnston at the same place and in the same month. S. anomalum is undoubtedly the one to which Saunders refers as 8. fasciculatum Thuret. It does not seem, how- ever, to be Thuret's species, if we have a correct conception of what S. fasciculatum Thuret includes. His species was published in Le Jolis, Algues marines de Cherbourg, no. 100, and in Liste, 1863, p. 73. Thuret quotes as a synonym, 8. volubilis Pringsheim (Beitr. Morph. Meeresalg., p. 13, pi. 3, fig. B, read in 1862). Pringsheim 's figure of volubilis shows the gametangia fasciculately branched, and he states that this character is one of the chief characters of the genus Stre- blonema. This figure has been much quoted since. Pringsheim gives no measurements of the parts, neither does Thuret. Reinke (Algenfl., 1889, p. 41) recognizes Streblonema as a subgenus of Ectocarpus. He recognizes Pringsheim 's plant, quoting the fore- going figure, but since volubilis was already occupied by Crouan (1867, p. 161) and fasciculatum was occupied by Harvey (Phyc. Brit,, pi. 273), he renamed the plant, calling it Pringshcimii. Hauck (1884, p. 323) seems to have been the first to give measurements of the various parts of the plant. He lists it under Streblonema, cites the above literature of Thuret, and quotes S. volubilis Pringsheim. Con- sidering Pringsheim 's plant, recognized by Reinke and Hauck, as being the same as Thuret's, and taking Pringsheim 's figure and Hauck 's measurements as being correct for Thuret's 8. fasciculatum, then our plant is distinct and undescribed. There is a little doubt in our minds at present whether the plant we have described as Streblonema anomalum is one or two species. We do not find the filaments bearing the zoosporangia in the type material examined to be like those figured by Saunders. His figures show the main filaments as being like those of all the known Streb- lonemas, monosiphonous, while those in both collections of material which we have examined are uniformly polysiphonous, that is, the cells of the main central parts of the thallus are divided once or twice lengthwise. This never takes place in the plants which bear game- tangia and which are intimately associated with them. The measurements, method of branching of the two sets of plants, and their chromatophore characters are practically the same. At least two interpretations of this polysiphonous phenomenon appeal 444 University of California Publications in Botany L VoL - 8 to us. They may represent a polymorphic state, in which a non-sexual plant differs in form from a sexual plant of the same species, a unique condition in the genus Streblonema, or the polysiphonous condition of the main filaments, found only in the non-sexual plants, may repre- sent a character belonging to a wholly different genus, and hence it is a new species of that genus. Until a more extensive study of fresh material can be made, we feel that it is best to take the former view, and place it in a new and polymorphic species of Streblonema. We have amended the family Ectocarpaceae to include species with this polysiphonous condition. 3. Streblonema Johnstonae S. and G. Plate 52, fig. 4 Fronds microscopic ; creeping filaments moderately branched, branches alternate or opposite, at times slightly secund ; erect filaments mostly simple, tapering slightly upward and at the base, extending beyond the surface of the host, in part short-pilif erous ; cells of the main creeping filaments more or less irregular in shape, chiefly doli- iform, 12-18/a diam., 1.5-5 times as long ; cells of erect filaments up to 24/A diam. in the widest part, nearly cylindrical, constricted at the cross-walls; chromatophore a thin parietal band, nearly covering the cell ; zoosporangia unknown ; gametangia cone-shaped to unsym- metrically fusiform, 90-130^ long, 28-36/t broad, mostly short-pedicel- late on the creeping filaments extending to the surface of the host. Growing in Cumagloia AnolersonU (Farlow) S. and G. in company with Streblonema corymbiferum and 8. anotnalum. San Pedro, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 394, pi. 43, fig. 4. Of the three plants found ramifying among the filaments of the above mentioned host, Streblonema Johnstonae is the most robust in all of its parts. It can readily be distinguished from the other two species with which it is associated by the large size of the gametangia. No zoosporangia have been observed. 4. Streblonema vorax S. and G. Fronds microscopic, prostrate filaments profusely branched, very tortuous, penetrating among the epidermal cells and spreading into the interior of the host, decomposing the walls and filling the large 1^25] Setchell-Gardner: Melcmophyceae 445 parenchyma cells and devouring their contents; cells 6-10/x long, 5— 8p diam.; erect filaments 400-600/x Long, 6.5-8/m wide, attenuated above into a hair, fasciculately branched at or near the surface of the host. forming a compact mass of cells; zoosporangia broadly elavate, 60-100/a long, 15-30/x, broad; gametangia numerous, lateral on short pedicels or sessile, narrowly cylindrical, blunt, 40-70/i. (up to 100/x) long, 7-9/* diam. ; loculi mostly uniseriate. Growing on the outer ends of the leaves of eel-grass, in the lower littoral and upper sublittoral belts. Sitka. Alaska. Setchell and Gardner, Phyc. Cont. V. 1922, p. 389, pi. 44. figs. 1, 2. . Of all the species of Streblonema thus far discovered on our coast, 8. vorax is the most destructive to the host. Although the penetrating filaments have abundance of chromatophores, the plants seem to be in a large degree parasitic. The cells of the host are closely compacted, and have thick walls, yet they are devoured in large quantities. We suggest the possibility of the secretion of enzymes with digestive power which act upon the cell-wall and its protoplasm, after which they are absorbed. This plant is found in company with several other small Melanophyceae, Chlorophyceae and Rhodophyceae. none of which penetrate the host. 5. Streblonema Porphyrae S. and G. Fronds mostly endophytic ; prostrate filaments very tortuous, pene- trating the cell-walls of the host between the cells, freely branched ; erect filaments pushing slightly beyond the surface of the host, occa- sionally branched ; hair filaments unknown ; cells of the creeping fila- ments 3-4/t diam., irregular in shape ; zoosporangia unknown ; game- tangia terminal in erect filaments, extending slightly beyond the sur- face of the host, fusiform to irregular in shape, 25-35/i. long, 5-8/* broad. Growing in the parenchymatous base of Porphyra naiadum on eel-grass. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 387, pi. 44, fig. 6. Streblonema Porphyrae is an exceedingly diminutive, though apparently distinct, species, confined, so far as we know at present, to the cushion-like bases of Porphyra naiadum. It ramifies deep into the tissue, apparently not entering the cells. 446 University of California Publications in Botany [Vol. 8 6. Streblonema penetrale S. and G. Fronds forming a continuous pulvinate stratum of indefinite shape and size on the stipe of the host; penetrating- portion composed of sparingly branched filaments extending relatively deep into the host and mostly perpendicular to its surface; erect filaments fasciculately branched at the surface of the host, 70-125/x long, tapering slightly upwards, not piliferous; cells of the penetrating filaments cylindrical to irregular, 6.5-8/* diam., 3-5 times as long; cells of the erect filaments cylindrical to slightly doliiform; 6.5-8/* diam., 1.5-2.5 times as long; chromatophores band-shaped ; zoosporangia unknown ; gametangia cylindrical to blunt-fusiform, more or less irregular, 30-40//. long, 8-11/* broad ; loculi uniseriate. Growing on the stipes of Hesperophycus Harvey anus. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 388, pi. 44, figs. 3, 4. Streblonema penetrale, like S. myrionematoides below, is a difficult species to classify. In its method of development and general struc- ture it approximates S. myrionematoides closely. The penetrating portion is much more highly developed than in that species, extending into the host among the cells to a depth of four or five times as great as the part which extends beyond the surface. The size of all the parts is, in general, greater than in S. myrionematoides. No horizontal filaments are present. The penetrating filaments are mostly per- pendicular to the surface. 7. Streblonema transfixum S. and G. Fronds forming patches 5-8 mm. diam. ; creeping filaments pene- trating deeply among the cells of the host, distorted, irregularly branched, 4-5/* diam., giving rise to scattered, short, erect, unbranched filaments protruding from the surface and mostly terminated by game- tangia; cells of the erect filaments cylindrical, 7-9/* diam., 0.75-1.5 times as long ; chromatophores band-shaped, nearly covering the cell wall ; zoosporangia unknown ; gametangia cylindrico-conical, blunt, 40-60,/* long, 8-12/* broad; loculi 1-2 seriate. Growing on Desmarestia herbacea. San Pedro, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 391. Streblonema transfixum forms definite, discolored areas on the surface of the host, readily detectable. The species was described from L925] Setchell-Gardner: Melanophyceae 447 dried material. The colls of the host were much collapsed and did not straighten out when soaked and boiled, hence the details of the species were not so thoroughly worked out as is desirable. The plants penetrate among the cells of the host and seemingly pass all the way through it. Just how much constitutes a single plant cannot at present be made out. 8. Streblonema irregulare Saunders Fronds consisting of irregularly branching prostrate filaments, applied closely to the host plant, from which arise numerous simple or sparingly branched erect filaments 1-2 mm. high, 9-14/* diam., with cells as long or twice as long as broad ; rhizoidal filaments arise from the lower surface of the prostrate filaments and penetrate the host, 10-14/x diam.. longer or shorter than the diameter; zoosporangia unknown; gametangia narrowly ellipsoidal, terminal or lateral on the erect filaments, 55-70ju long, 14-18/*, diam. Growing on the cysts of N&reocystis Luetkeana. Sitka, Alaska. Saunders. Alg. Harriman Exp., 1901, p. 417, pi. 45, fig. 2; Setchell and Gardner. Alg. N.W. Amer, 1903, p. 239. We have not had an opportunity to examine the type material of this species, nor have we seen any other material representing the species since its discovery and first publication. 9. Streblonema minutissimum Saunders Fronds composed of short, sparsely branched, creeping filaments, 1-2//, diam., ramifying among the cortical filaments of the host, some- times applied to them, and of erect filaments arising from the prostrate filaments in the host, and intermingling with its peripheral filaments; erect filaments once or twice dichotomously branched, bearing above a few short branches which become transformed into uniseriate game- tangia, 20-30^1 long, 3-5/x broad. "Growing in the branches of Liebmannia.' n Sitka, Alaska. Saunders, Alg. Harriman Exp., 1901, p. 416, pi. 45, fig. 3 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 239. Nothing further has been learned of this species of Streblonema nor of the host upon which it grew since its discovery at Sitka and publication by Saunders (loo. eit.). We have not seen any of the specimens, but. judging from the figure and deseription, it seems distinct from all other known forms. 4 IS University of California Publications in Botany L VoL - 10. Streblonema pacificum Saunders (Orthog. mut.) Plate 41, fig. 53 Fronds occupying circular areas, 2-4 mm. diam. ; composed of a network of profusely and irregularly branched horizontal creeping filaments, penetrating among the paraphyses and sporangia of the host, giving rise at right angles below to numerous acute branches, 3-4 cells long, and above to numerous fasciculately branched filaments extending to the surface of the host and bearing the gametangia ; cells of the horizontal filaments 3.5-4/t diam., somewhat irregular ; cells of the rhizoidal filaments 1-2.5/* diam.; cells of the erect filaments variable in shape and size, 4-6/* diam. ; zoosporangia unknown ; game- tangia numerous, terminating the erect filaments and projecting beyond the surface of the host, blunt fusiform, 20-28/* long, 4-6/* broad ; loculi uniseriate. Growing on the sporophylls of Alaria, Alaska (Yakutat Bay) to California (San Francisco Bay). "Streblonema pacifiea" Saunders, Alg. Harriman Exp., 1901, p. 417, pi. 45, figs, la, lb ; Setchell and Gardner, Alg. N.W. Amer., 1903. p. 239. Superficially Streblonema pacificum Saunders appears like a small Myrionema, the plants forming circular masses in the same manner as that genus. The plants are exceedingly difficult to interpret in the stage in which they begin to be visible to the unaided eye, that is when the gametangia are forming. The sporangia of the host are at this time elongating and more or less disturbing the arrangement of the filaments. As far as can be ascertained from the study of mature and nearly mature plants, the circular masses are occupied by a num- ber of individuals instead of a single individual as in the case of a Myrionema. This may not be the case, however. Against this con- clusion the argument might be urged that the groups of plants are too nearly uniform in size and too nearly circular in outline to be composed of a necessarily variable number of plants intermingled, yet this is the way they appear at maturity. More study of the early stages in the life-history will have to be resorted to before the question can be settled. Saunders {Joe. cit.) figures hair filaments but does not mention them in his description. We have not been able to find any true hairs in the material from California, nor in the portion of the type which we examined. Neither have we been able to find any anastomosing of 1925] Setchell Gardner: Melamophyceat 44!) filaments such as he represents in figure lb (Joe. rit.). These latter conditions we may have overlooked in onr study, as the plants are exceedingly difficult to trace. The hair character is one that varies considerably in any species. 11. Streblonema evagatum 8. and G. Fronds forming circular masses 1-2 cm. diam. ; creeping filaments irregular, much branched, penetrating among the sporangia of the host at their bases; erect filaments fasciculately branched, usually within the host, tapering very slightly at the apex and at the base. 190-230/* high ; cells of the creeping portion 3.5-4/x diam. ; cells of the erect filaments 4,u diam. at the base, 2-3.5 times as long, 6.5-7.5/* diam. in the widest part, 1-2 times as long, cylindrical, slightly con- stricted at the cross-walls; chromatophores band-shaped, 1-2 in a cell; zoosporangia unknown; gametangia numerous, cylindrical, lateral, sessile or on short pedicels, 65-80jm long, 5.5-6.5/* diam. ; loculi uniseriate. Growing in the blades of Lam nutria Andersowii, upper sublittoral belt. Cypress Point, Monterey County, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 390, pi. 42, figs. 1-5. This species has a habit of growth very similar to that of Streb- lonema pacificum, but the plants cover much more extensive areas which are quite certainly not infested by a single plant as may be the case in that species. Their presence seems to affect the host materially, to a large extent destroying its sporangia and causing discoloration. It is possibly partially parasitic. 12. Streblonema rugosum S. and G. Plate 52, figs. 5-7 Fronds microscopic, producing decidedly rugose areas of consid- erable extent on the surface of the host ; creeping portion penetrating among the cells near the surface of the host, giving rise above to fasciculately branched, erect, fructiferous filaments and hair fila- ments; cells of creeping and erect filaments very irregular in shape and size; chromatophores band-shaped; zoosporangia unknown; game- tangia blunt, fusiform, terminal on most of the erect filaments, 18-24/* long, 5-6.5/* broad ; loculi uniseriate, 4—6 in a series. Growing on the blade of Alaria temiifoHa, Friday Harbor, Washington. Setchell and Gardner, Phyc. Cont. V, 1922, p. 390, pi. 43, figs. 5-7. 450 Unwersity of California Publications in Botany L v ol. 8 This species, like Streblonema my rionemat aides and S. scabiosum, has the larger part outside of the host. It penetrates to a considerable depth among the cells of the uninjured host, later apparently mechanic- ally causing the death of a few surface cells and an abnormal growth of others, giving to the surface a rugose appearance. The plants seem to associate in small groups of indefinite shape and size, often confluent and although small, can readily be detected by the peculiar appearance they give to the host. 13. Streblonema scabiosum S. and G. Fronds forming circular to elliptical pustules of indefinite size on the host ; penetrating portion profusely branched, at first pushing between the cells causing their death, later taking possession of the lumen of the cell, causing complete disintegration ; erect filaments extending beyond the surface of the host, unbranched or more or less f asciculately branched at the base, 50-80/* high ; cells of the creeping filaments very irregular in shape and size ; cells of the erect filaments cylindrical, 4.5-5.5/* diam., 1-2.5 times as long; chromatophores band- shaped; zoosporangia unknown; gametangia cylindrical, sessile or short-pedicellate, extending beyond the surface of the host, 40-60,u. long. 4.5-6/* broad. Growing on the stipe of Nereocystis Luetkeana. Cast ashore near the "Cliff House," San Francisco, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 389, pi. 44, fig. 5. Streblonema scabiosum forms marked scrofulous-like areas on the lower part of the stipe of the host. It has a disastrous effect upon the host, destroying its cells as far as it penetrates, and seems con- stantly to spread by new infections around the margin of the affected area. The cause of the death of the cells of the host has not been investigated. In habit of growth and general form the species seems related to 8. myrionematoicles and to 8. penetrate. 14. Streblonema aecidioides f. pacificum S. and G. Fronds microscopic, showing on the surface of the host as small elevations, 75-150/* diam., the vegetative part forming a more or less parenchymatous layer just beneath the surface layer of cells of the host and giving rise below to a few rhizoidal filaments penetrating deeper into the host ; all erect filaments fructiferous except a few short hair filaments arising in the center of the frond; cells of the hair filaments 4-5.5/* diam., quadrate below, 5-8 times as long above, i 9 - 5 ] Setchell-Gardner: Melanophyceae 451 without a sheath; zoosporangia(?) narrowly clavate, sessile, 22-28/u long, 8-12/x broad at the apex ; gametangia numerous, closely crowded, cylindrical, sessile on the horizontal layer, 45-55^ long, 5-6.5^ broad; loculi uniseriate. Growing within the lamina of Hedophyllum sessile (Aresch.) Setchell, near the outer end. Neah Bay, Washington. Setchell and Gardner, Phyc. Gout. V, 1922, p. 395, pi. 44, figs. 8, 9. Phycocelis fecunda Tilden, Amer. Alg. (Exsicc), no. 356 from Victoria, B. C. (determined by De Alton Saunders). Streblonema aecMioidt s f. pacific ion seems very closely related to the Ectocarpus aecidioides of Rosenvinge (1893, p. 894), found grow- ing in Greenland on Lammaria longicruris and L. groenlandica. It differs only in minor details as to the dimensions of the parts. Regard- ing the zoosporangia we have to speak with uncertainty. As figured by Rosenvinge, the Greenland plant has them well developed and producing zoospores. They are in distinct "aeeidia" and apparently on distinct non-sexual plants. In our species they likewise appear to be on non-sexual plants. In ours there is no indication of the produc- tion of zoospores. The plants are possibly too young, or possibly they bear abortive organs so commonly met with in various other genera on our coast, the nature of which is still an open question. This form seems to enter the host from the surface and after penetrating to the second layer of cells spreads out horizontally between the sur- face layer and the second layer. Later, filaments arise from the under side of this layer and penetrate among the cells of the host, apparently never entering them. Finally from the upper surface each cell in the central region of the layer gives rise to a filament and the mass acting together lifts up the surface layer of cells of the host, forming a small blister which finally ruptures, as in the case of Rosen- vinge 's plants. The erect filaments are almost simultaneously trans- formed into gametangia, except a few in the center which develop into hairs. Foslie (1894, p. 167, 23 in reprint) describes forms of the same species found growing on Lammaria saccharina at Kjelvik and at Lyngd near Troms0, Norway. These he listed under Ectocarpus (Streblonema) aecidioides Rosenvinge. The measurements of the parts of his forms average, in general, a little larger than those of Rosenvinge. De-Toni (1895, p. 577) cites these Arctic plants under Streblonema aecidioides (Rosenv.) Foslie. Foslie gives 8(ht as the extreme length of the gametangia. 452 University of California Publications in Botany [Vol. 8 15. Streblonema myrionematoides S. and G. Fronds microscopic ; prostrate portion very poorly or scantily developed, penetrating only slightly among the two or three outer layers of cells of the host ; erect filaments more or less f asciculately branched at the surface of the host, 65-80//, long, mostly fructiferous, hairs few ; cells of the penetrating filaments 4-5/i. diam., irregular in shape ; zoosporangia unknown ; gametangia numerous, cylindrical, 50-65/x long, 4.5-6. 5/a broad ; loculi uniseriate. Growing on the blade of Laminaria Andersonii Farlow. Moss Beach, San Mateo County, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 387, pi. 44, fig. 7. It is an extremely perplexing problem to decide upon the generic position of forms such as the one described above. It has close affini- ties with the three genera, Streblonema, Ectocarpus and Myrionema. The erect fronds exterior to the host and more or less branched, with relatively scanty attaching portions, constitute the vegetative portion of an Ectocarpus of microscopic size. The gametangia are strictly those of a typical Myrionema, but there is lacking the disk-shaped basal layer of filaments spreading out on the surface of the host, character- istic of that genus. The extremely reduced character of the penetrat- ing portion does not speak well for the genus Streblonema. On the whole, we feel that with our present criteria for these genera it best agrees with the characters of the genus Streblonema, where we have placed it. It penetrates the uninjured host, but only to a slight depth. The plants, however, are usually so congested that their growth soon crowds the surface layer of the host cells to such an extent that they die and disintegrate, there being no evidence that the associate pene- trates them and absorbs their material. The palisade-like stratum of gametangia suggests very strongly the Myrionema character which is the reason for the specific name. 16. Streblonema investiens (Collins) S. and G. Fronds occupying indefinite areas on the host ; creeping filaments irregularly branched, often curving outward and bearing on the out- side short, simple, or sparsely branched filaments; hairs sparse; cells of creeping filaments 5-8/x diam., 1-2 (up to 4) times as long, swollen or cylindrical; cells of the ramuli 6/x diam., 1-2 times as long; cells of the hairs 8/x diam. ; chromatophores discoid, small, several in a 1925] Setchellr-Gardner: M< lanophyeeae 453 cell; zoosporangia ovoid, sessile or on 1-celled pedicels on both the creeping filaments and the ramuli, 20/a long-, 15/*. broad; gametangia cylindrical. 25-40/* long, 8-10^ broad ; loculi mostly aniseriate ; game- tangia and zoosporangia growing on the same plant. Growing in the fronds of HeLminthochtdia calvadosii (Lamour.) Setchell. San Pedro, California. Setchell and Gardner, Phyc. Cont. V, 1922, p. 396. Strepsithalia investiens Collins, in Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 738. It seems that the chief technical distinction between the genera Strepsithalia Sauvageau and Streblonema Derb. and Sol. is the secre- tion by Strepsithalia of a rather copious gelatinous sheath investing the entire plant, particularly the exposed portions, the ramuli. Since we are not able to demonstrate the presence of such a sheath, even to the slightest degree, we have thought it best to place Collins' Strepsithalia investiens in the genus Streblonema. family 3. MYRIONEMATACEAE foslif. (orthog. mut.) Thallus two-fold, viz., (1) of a prostrate disk formed of radiating filaments more or less closely united and one or two cells in thick- ness, entirely superficial or penetrating- the host by short rhizoidal filaments, and (2) of erect, simple or branched, monosiphonous, usually parallel filaments, either free or enclosed in a common jelly ; growth peripheral in the basal portion and subapical or intercalary in the erect filaments ; hairs present or absent ; erect filaments entirely fertile, partially sterile or sterile and fertile intermixed; cells uninu- cleate, each with one or more parietal disk-shaped or band-shaped chromatophores ; zoosporangia unilocular, terminal or lateral, sessile or pedicelled, never intercalary; gametangia plurilocular, terminal or lateral, with loculi uni- or pluriseriate; gametophyte and sporophyte identical in size and structure. My Hone mac eae Foslie, Contr. Knowl. Mar. Algae Norway, I, 1890, p. 91. The members of the Myrionemataceae, as we understand them. seem closely related to the Ectocarpaceae, on the one hand, and the Elachisteaceae on the other. There seems to be little reason for plac- ing the genera in Chordariaceae as Kjellman (1893, p. 225) has done, nor even associating them closely with the Mesogloiaceae, as Oltmanns (1922, p. 29) has done, since it seems to us that Strepsithalia, in typical form, resembles Streblonema in its basal portion, but agrees 4.14 University of California Publications in Botcmy [Vol. 8 better with the Elachistea and Myriactis-groxrps of the Elachisteaceae in its erect filaments slightly contracted at the base. The various .Myrionemataceae are most closely allied to the species of Ralfsia, particularly with those referred to Stragularia. The prostrate basal portion of one or two layers of cells readily distinguishes the members of the Myrionemataceae from those of all other families except those of the simpler Ralfsiaceae. Key to the Genera 1. Erect filaments free 2 1. Erect filaments enclosed in a jelly 9. Microspongium (p. 492) 2. Basal portion monostromatic 3 2. Basal portion distromatic 8. Hecatonema (p. 488) 3. Gametangia with loculi uniseriate 6. Myrionema (p. 454) 3. Gametangia with loculi pluriseriate 7. Compsonema (p. 473) 6. Myrionema Grev. Fronds forming small cushions, circular to ellipsoidal, or even quite irregular in outline, composed of a monostromatic basal stratum and numerous pigment-bearing erect filaments with or without hairs interspersed; the erect filaments in part or wholly transformed into gametangia, except at the margins ; basal stratum composed of closely crowded filaments radiating from a common center, with dichotomous branching by splitting of the terminal cells, rarely with a few, short, subulate branches penetrating the host; reproduction by unilocular zoosporangia and by plurilocular gametangia mostly with uniseriate loculi. Greville, Scottish Crypt. Flora, vol. 5, 1827, pi. 300. The type species of this genus is M. strangukms Grev. The type locality is the shore of Appin, Scotland. It was first brought to notice by Captain Carmichael, who found it growing on a small species of Solenia (Enteromorpha), which it usually completely encircles. Greville says of his genus : "It evidently belongs to the Nostochinae of Agardh." Although it is undoubtedly known today to belong to the Melanophyceae, there has been considerable diversity of opinion concerning the family to which it properly has its closest affinities. Harvey (1841, p. 124) placed it among the Chlorospermeae, under the tribe Chaetophoroideae, probably on account of its "gelatinous" nature as suggested by Greville. He states, however, that it is "a genus of doubtful affinity," and "it rather belongs to the series Melanospermeae. " J. G. Agardh (1848, p. 47) placed it in his order Chordarieae under his tribe Mesogloiaceae. Kjellman (1890, p. 40) 1923] Setchellr-Gardntr: Melanophycecu 4.").") emended the genus and placed it in the family Chordariaceae, and in 1893 (p. 226) retained it in the same family. Foslie (1S!)(). p. 91) and Rosenvinge (1893, p. 901) and si ill later, Borgesen (1902, p. 41 9 I placed it in the ''Myrionemaceae, " the first two designating the group as a family, and the last as an order. De-Toni (189.1, p. 17s i retained it in Chordariaceae. Oltmanns (1904, p. 382) placed it in the group Myrionemeae of the family Ectocarpaceae. In 1922 he arranged it under the "Isogame Ectocarpales" in the family "Myrionemaceae," along with Strepsithalia and Ralfski, After a careful study of a large assortment of material belonging to the Ectocarpales, we have limited the genus to include forms possessing the characters mentioned in the foregoing generic diagnosis and placed the genus in the family Myrionemataceae, along with three other very closely related genera; viz., Compsonema, Microspongwm, and Hecatonema, We have, however, included under JI>/rionema the species formerly assigned to AscocycLus Magnus and Phycocclis Stroemfelt. In working over the members of the genus Myrionema found on the Pacific Coast of North America, we have studied and separated over twenty species or forms. This segregation must be considered as provisional since no attempt at cultural criteria has been possible. We have separated the species into two lines, the one in which only zoosporangial forms are known, and the other in which only game- tangial specimens are known, or those having gametangia and " ascocj^sts, " the latter problematic structures. Possibly some of the structures we have called "ascocysts" may be young zoosporangia, but probably most or all of such structures, as we have seen them, are hypertrophied gametangia. The possible genetic connection between some of the one series of species and some of the other series cannot be demonstrated except by cultures. Key to the Species 1. With zoosporangia only 2 1. With gametangia only or with both gametangia and "ascocysts" 3 2. Plants forming definite circular thalli 21. M. strangulans (p. 471) 2. Plants forming indefinite rugose thalli 22. M. obscurum (p. 472) 3. Gametangia sessile 4 3. Gametangia pedicellate 9 4. All of the erect filaments transformed into gametangia 5 4. Part of the erect filaments transformed into gametangia 10. M. foecundum f. divergens (p. 463) 5. Plants with subulate rhizoids 7. M. foecundum f. subulatum (p. 462) 5. Plants without subulate rhizoids 6 6. Gametangia branched in part 8. M. foecundum f. ramulosum (p. 462) 6. Gametangia unbranched 7 4.">(i University of California Publication's in Botany [Vol.8 7. Gametangia up to 40/i long 6. M. foecundum f. simplicissimum (p. 461) 7. Gametangia over 40m long 8 8. Gametangia up to 65m long and 6m broad 1. M. primarium (p. 456) 8. Gametangia 80m long and 8m broad 9. M. foecundum f. majus (p. 463) 9. All of the erect filaments fructiferous 10 9. Part of the erect filaments fructiferous 12 10. Gametangia up to 130m long 12. M. balticum f. californicum (p. 465) 10. Gametangia up to 45m long 11 11. Pedicels 1-3-celled 3. M. corunnae f. uniforme (p. 458) 11. Pedicels up to 8-celled 11. M. minutissima (p. 464) 12. Sterile filaments longer than the gametangia 13 12. Sterile filaments not longer than the gametangia 20 13. Gametangia all terminal 14 13. Gametangia in part lateral on sterile filaments 19. M. setiferum (p. 470) 14. Sterile filaments partially setiferous 15 14. Sterile filaments not setiferous 16 15. Erect filaments unbranched 18. M. phyllophilum (p. 469) 15. Erect filaments branched 20. M. globosum f. affme (p. 471) 16. Gametangia decidedly acuminate 2. M. primarium f. acuminatum (p. 457) 16. Gametangia cylindrical or nearly so 17 17. Cells of sterile filaments doliiform 17. M. attenuatum f. doliiforme (p. 468) 17. Cells of sterile filaments cylindrical 18 18. Creeping filaments with short subulate rhizoids 15. M. hecatonematoides (p. 467) 18. Creeping filaments without rhizoids 19 19. Gametangia up to 45m long 16. M. attenuatum (p. 468) 19. Gametangia up to 150m long 13. M. balticum f. pedicellatum (p. 466) 20. Erect filaments up to 65m high 5. M. corunnae f. sterile (p. 460) 20. Erect filaments over 65m high 21 21. Gametangia on 1-3-celled pedicels 14. M. compsonematoides (p. 467) 21. Gametangia on long pedicels, 7-10-celled .. 4. M. corunnae f. angulatum fp. 459) 1. Myrionema primarium S. and G. Plate 55, fig. 12 Fronds forming circular cushions 0.5-1.25 mm. diam. ; prostrate portion composed of regularly radiating, closely compacted filaments ; erect filaments unbranched, densely crowded, arising from every cell of the creeping filaments; true hair filaments numerous, scattered promiscuously in the center of the frond ; cells of erect filaments 4— 4.5//. diam.. quadrate below, terminal cell, 2-3 times as long when young ; cells of true hairs 4-5//. diam., 4—6 times as long; zoosporangia unknown; gametangia sessile on the creeping filaments, arising by transformation of all of each erect filament, except at the margin, very compact, of nearly equal length in the center of the frond, shorter toward the margin, cylindrical, 55-65//. long, 5-6//. broad, blunt or tapering slightly at the apices, terminal cell acute-conical. 1925] Setchell-Gardner: Melanophyceae 457 Growing on the outer end of the young blade of Costaria costata in the lower littoral and the upper sublittoral belts. Oregon (Coos Bay) to central California. Setchell and Gardner, Phye. Cont. II, 1922, p. 334, pi. 34, fig. 12. We consider Myrionema primarium to be a typical representative of a group whose members are the most primitive of the genus when considered from the standpoint of differentiation. The prostrate basal layer is composed of long, regularly radiating filaments with apical growth. Radial divisions of apical cells occur just often enough to completely occupy all of the space between one another as the plant continues to increase in diameter and in circumference. The branch- ing is always dichotomous, and is accompanied by widening of the apical cell, and the establishment of two growing regions on opposite corners, which are subsequently separated from the remainder of the cell by a wall, thus establishing two equal branches. Beginning in the center, each cell successively toward the periphery gives rise by hori- zontal divisions to erect filaments of nearly equal length, the only exception or modification being that some cells give rise to long filaments, the so-called true hairs, with the meristem at the base, and the outer cells long and colorless. All other erect filaments are transformed into gametangia. Only the cells of the basal filaments remain sterile, or probably a few short erect filaments at the periphery of the frond may never come to maturity. Thus starting with a single cell the maximum of reproductive cells arises in this group. Starting with such forms the course of evolution seems to have been in the direction of the sterilization of tissue. 2. Myrionema primarium f. acuminatum S. and G. Fronds microscopic, growing among other small algae, erect fila- ments in part sterile, 80-100/t long, cells slightly doliiform ; game- tangia 4-5//. diam., acuminate, in part short-pedicellate. Growing on Macrocystis pyrifera. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 335, pi. 32, fig. 9. There is but a slight difference between M. primarium and forma acuminatum. It is worthy of note that many gametangia in the center of the thallus are short-pedicellate, and that a few, about one in twenty-five, of the erect filaments continue to grow, attaining a length two to three times as long as the gametangia. 458 University of California Publications in Botany [Vol.8 Myrionema corunnae Sam-. Fronds forming: small, circular cushions, 2-3 mm. diam., basal disk monostromatic, not attached by rhizoids from below; erect fila- ments longest in the middle of the thallus, gradually reduced in length toward the margin ; hairs scattered, relatively short, 4-5.5/* diam. at the base, cells 25-40/* long at the apex, meristem at the base but all cells elongating at maturity ; all erect filaments fructiferous ; game- tangia on 2-4-celled pedicels at the center of the frond, becoming sessile toward the margin, mostly simple but not infrequently bearing 3-5 branches, 25-120/1. long, 4.5-7/* diam. ; zoosporangia unknown. Sauvageau, Sur quelq. Myrionem., 1897, p. 237 (repr. 1898. p. 77). The type material of the species was found growing in minute, closely associated cushions on the outer ends of Laminaria pallida at La Corogne (Spain), in November, 1895. We have not yet dis- covered in our region any specimen which we can refer to typical M. corunnae as described and figured by Sauvageau. The material collected by one of us (Gardner) in December, 1909, growing on Nereocystis Luetkeana (Mert.) P. and R., was identified by F. S. Collins as belonging to that species and distributed in Phycotheca Boreali-Americana (Exsicc.) as no 1738. Careful study of this material reveals a mixture of several genera belonging to Myrione- mataceae and Ectocarpaeeae, none of which, so far as the specimens in our copies go, is 31. corunnae. Experience in collecting and critical study of the numerous forms found growing upon the above mentioned host, have shown the necessity of exercising extreme care in selecting pure or uniform material. The presence of branched gametangia in typical M . corunnae, as diagnosed by Sauvageau, is a character not found in any of our Pacific coast forms. ' The three following distinct forms have been segregated : 3. Myrionema corunnae f. uniforme Fronds forming irregular cushions, 0.5-1.5 mm. diam. ; creeping filaments forming a compact layer with peripheral growth ; hairs absent ; erect filaments unbranched, slightly attenuated at the apices, 80-100/* (up to 120/*) high; cells of erect filaments 4.5-5. 5/* diam.. 1-1.5 times as long; "ascocysts" sparse, clavate; gametangia forming a solid, compact, palisade stratum of approximately the same height 1925] Setckell-Gardner: Melanophyceae 459 in the center, gradually diminishing toward the periphery, mostly sessile, some on 1-3-eelled pedicels, 80-10G> high, 6-6. 5/t diam. ; loculi uniseriate. Growing on the blades of Costaria costata, Laminaria SincJairii and Alceria marginatu. Central California (Moss Beach, San Mateo County, and Fort Point in San Francisco). Setchell and Gardner, Phyc. Cont. II, 1922, p. 339, pi. 33, figs. 4, 5. Forma uniforme differs from the species as described and figured by Sauvageau (1898, pp. 77-82, repr., figs. 14 A-F) in having no hair filaments, in having no branched gametangia, in having "ascocysts," and in having slightly different dimensions. There are some differences in the size of the plants we have found growing on different hosts in central California. The plants chosen as the type of the form grew on the blades of Alaria marginata at Fort Point, San Francisco. Very generally, this species of Alaria has delicate grooves obliquely radiating from the midrib to the margin. The presence of these grooves is often accentuated by the growth of this minute Myrionema, usually so numerous as to be confluent, so that the individuals cannot be detected except by microscopic examination. 4. Myrionema corunnae f. angulatum S. and G. Fronds forming microscopic cushions, irregular in outline, among other small algae on the host ; creeping filaments regularly radiating when undisturbed ; erect filaments unbranched, densely crowded, 120-140/a long; true hairs absent; cells of creeping filaments 4-4.5/t diam., mostly quadrate; cells of erect filaments cylindrical, 4-4.5//. diam., 1.5-3 times as long; zoosporangia and "ascocysts" unknown; gametangia cylindrical, slightly attenuated at the apices, terminal on 6-10-celled pedicels, 35-45/* (up to 6Gy) long, 4-6.5/x broad, many partitions of the loculi oblique. Growing on the cysts of Egregia Menziesii. Cast ashore near the ' ' Cliff House, ' ' San Francisco, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 340, pi. 33, figs. 6, 7. This form is to be distinguished from the species as described by Sauvageau (1898, pp. 77-82, repr.) by the absence of hairs, the presence of sterile erect filaments, the gametangia mostly on longer pedicels and by the very small size of the loculi, often produced by perpendicular and slanting walls. In part, the gametangia seem pluri- seriate, caused by the division of some of the original cells into small, angular loculi. 460 University of California Publications in Botany [Vol. 8 5. Myrionema corunnae f. sterile S. and G. Fronds forming circular cushions, 1.5-3 mm. diam. ; erect filaments straight unbranched, cylindrical, blunt, 55-65/x long; true hairs and "ascocysts" absent; gametangia cylindrical, blunt, on 3-5-celled pedicels, rarely sessile, 55-65/a long, 4.5-5.5/* broad; gametangia and sterile erect filaments diminishing in length very regularly and gradually from the center to the circumference of the fronds. Growing on the pneumatocysts of Nereocystis Luetkeana. Cast ashore. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 340, pi. 33, figs. 13-15. Chilionema ocellatum Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1739. This form is to be distinguished from the species as described by Sauvageau (1898, pp. 77-82, repr.) by the presence of numerous sterile erect filaments interspersed among the gametangia, about one- fourth to one-fifth as many as the gametangia, by the absence of hairs, and by the very gradual diminution in the length of the sterile fila- ments and the gametangia from the center to the circumference of the fronds. The sterile erect filaments are of the same length as the gametangia. Myrionema foecundum (Stroemf.) Foslie (Orthog. mut.) Myrionema (Phycocelis) foecunda (Stroemf.) Foslie, New or Crit. Norw. Algae, 1894, p. 17 (repr.). Phycocelis foecunda Stroem- felt, in Notarisia, 1888, p. 383, pi. 3, fig. 5. Foslie (1894, p. 15, repr.) states, ''I have been unable to find any real distinction between Myrionema, Phycocelis and Ascocyclus, and therefore I consider the two latter as subgenus of the former, Ascocyclus, however principally founded on the presence of para- physes. " We likewise are unable to discern any fundamental dis- tinctions among these three genera. Phycocelis was proposed by Stroemf elt (1888, p. 383) with P. foecunda as the type. This species has a unistratose base, and all of the erect filaments, except the hairs and the marginal filaments, metamorphose into gametangia. These are characters of typical Myrionema, as we interpret the genus. The genus Ascocyclus was founded by Magnus (1874, p. 73) based upon J. G. Agardh's Myrionema orbiculare, which, as figured by Crouan (1867, pi. 25, gen. 163), has a unistratose base, plurilocular 1925] Setch< ll-Oardm r: Melanophyceas 461 gametangia and hairs. Having found in addition to these certain single celled sack-like hairs, he created the new genus based upon this character. Two sizes of these are well illustrated by Kylin (1907, p. 39, fig. 9), found growing on Zostera and Cladophora. These are probably the structures which Foslie (Joe. cil.) refers to as "para- physes. " After having studied these structures as they occur in a variety of species of different genera belonging to the Ectocarpales, our experience leads us to conclude that they are not constant in any species, and that they are probably caused by some pathological con- dition, the nature of which we are not at present able to state. We are taking the view, therefore, that neither of the two genera, Phycocelis Stroemfelt and Ascocyclus Magnus are valid genera, or even subgenera. We have not seen either the type or any other authentic material of the Phycocelis foecunda Stroemfelt, neither have we any authentic measurements of the original material. De-Toni (1895, p. 582) gives 35/x to 70fi as the height and 7/x to 12/a as the width of the gametangia. We are listing five forms under the species none of which, so far as we are able to judge, corresponds to the typical material. 6. Myrionema foecundum f. simplicissimum 8. and G. Fronds forming circular cushions 0.5-1.5 mm. diam. ; prostrate portion composed of regularly radiating, closely adhering filaments with apical growth and branching; erect filaments all fructiferous; true hairs unknown ; cells of creeping filaments 3-4.5/* diam., quadrate toward the center, two times as long at the margin ; zoosporangia unknown; "ascocysts" rare to numerous, clavate, 28-34/;. high; game- tangia arising by transformation of all of each erect filament, except near the margin of the frond, 30-40//, high, 7-9//, broad. Growing on the sporophylls of Nereocystis Luetkeana. Alaska (Kadiak Island), to central California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 336, pi. 33, figs. 9, 10. Myrionema strangulans Setchell and Gardner, Alg. N.W. Amer., 1903, p. 249; Myrionema vulgare Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 924. Phycocelis reptans Tilden, Amer. Alg. (Exsicc), no. 357 (in part). The Phycocelis reptans of Tilden, mentioned above, proves not to be uniform, so far as the distribution in the copy at hand is concerned. The material growing on the sporophyll of Nereocystis seems clearly to be M. foecundum f. simplicissimum although young. The material 462 University of California Publications in Botani/ [Vol.8 on the other part of the host is a mixture. In height of gametangia forma simplicissimum approximates to forma seriata Reinke (1889, p. 46), but has no true hairs so far as we have been able to observe from the examination of many specimens from different localities. At times the individuals are very numerous on the host, so much so as to interfere with each others' growth. In one collection (Gardner, no. 4651) they are completely confluent over several square inches, and the outline of the individuals can be detected only by staining and by microscopic examination. 7. Myrionema foecundum f. subulatum S. and G. Basal filaments attached by numerous awl-shaped rhizoicls, 2-4 cells long, penetrating between the cells of the host; zoosporangia, "ascocysts" and true hairs unknown. Growing on the sporophylls of Nereocystis Luetkeana. Near the ' ' Cliff House, ' ' San Francisco, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 337, pi. 33, figs. 1-3. Forma subulatum is very similar to forma simplicissimum. The gametangia average slightly larger and are usually more blunt. There are no so-called "ascocysts" present. The chief distinction is the presence in this form of numerous awl-shaped rhizoids, found in no other form of the species, so far as we have been able to ascertain. 8. Myrionema foecundum f. ramulosum S. and G. Fronds forming small circular cushions, 1-3 mm. diam. ; erect filaments slightly branched above, 55-65/x long ; cells of the erect fila- ments 5.5-6.5/* diam., quadrate, terminal cell frequently 2-3 times as long; zoosporangia, "ascocysts" and true hairs unknown; all of the erect filaments transformed into gametangia. Growing on the pneumatocysts of Nereocystis Luetkeana. Friday Harbor, San Juan Island, Washington. Setchell and Gardner, Phyc. Cont. II, 1922, p. 337, pi. 33, fig. 8. This form of 31. foecundum differs from the other forms of the species in having secund "proliferations" near the outer ends of many of the gametangia. i !, --5] Setchell-Gardner: Melanophyceae 4(i:i 9. Myrionema foecundum f. majus S. and G. Fronds forming circular cushions, 1-2 mm. diam. with convex upper surface; prostrate portion composed of closely adhering, regu- larly radiating filaments ; erect filaments densely crowded, unbranched ; cells cylindrical, 5-7/i diam. ; true hairs sparse, distributed promiscu- ously over the frond, meristem at the base, surrounded by a sheath, 400-500/x long, cells 6/x diam., 5-8 times as long as the diameter at the apices; "ascocysts" sessile on the prostrate filaments, cylindrical, slightly larger at the distal end, 45-55/a long, 8-12/a diam., scattered sparsely over the frond, occasionally in groups; zoosporangia un- known ; gametangia sessile on the basal filaments, cylindrical, 60-80/x long, 6-S/x diam., arising by transformation of all of each erect filament, cross- walls horizontal or oblique. Growing on Laminaria ephemera Setchell. Port Renfrew, Van- couver Island, Neah Bay, Puget Sound and central California (Carmel Bay). Setchell and Gardner, Phyc. Cont. IT, 1922, p. 338, pi. 32, fig. 7. This small epiphyte seems to be restricted in its habitat to the above named host. The only known localities in which the host has been observed are those mentioned above. The material from each locality is richly infested by this epiphyte. The host, richly infested, was distributed by Miss Josephine Tilden in her American Algae (Exsicc), no. 609, under Benfrewia parvida Griggs. It seems very closely related to Myrionema foecundum f. simpli- cissimum. The differences are in the larger measurements of all of the parts, the presence of numerous, larger, so-called "ascocysts," occurring in groups, or scattered among the gametangia, in the presence of well-developed hairs, and in the more convex upper surface of the mass of erect filaments. 10. Myrionema foecundum f. divergens S. and G. Plate 55, figs. 8, 9 Fronds forming circular cushions, 1.5-3.5 mm. diam. ; basal por- tion composed of regularly radiating filaments closely compacted, firmly adhering to each other and to the host ; sterile erect filaments sparse, unbranched, blunt, scattered promiscuously among the game- tangia, 300-400^ long; true hairs absent; "ascocysts" numerous, scattered among and shorter than the gametangia, cylindrical to clavate; cells of the creeping filaments 5-6/* diam., 1.5-2 times as 464 University of California Publications in Botany ITol. 8 long; cells of the erect filaments 8-fy* diam., 1.5-3 times as long, cylindrical, not constricted; zoosporangia unknown; gametangia variable in shape, cylindrical, clindri co-conical or fusiform, 50-60/* long, 7.5-8.5/* broad; loculi 1-2-seriate. Growing on the sporophylls of Nereocystis Luetkeana. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 338, pi. 34, figs. 8, 9. This form of M. foecundum manifests some slight differentiations not present in any other form of the species. Scattered promiscuously over the fronds are erect filaments two to three times as long as the gametangia. The "ascocysts" are abundant. Some irregularities are also shown in the creeping filaments. Typically the branching: seems to take place by the splitting of the apical cell, but it seems that frequently one of the dichotomy fails to develop till later, pro- ducing the appearance of lateral branching. There is also an occa- sional oblique or longitudinal division of a cell of the creeping fila- ments perpendicular to the host. 11. Myrionema minutissimum S. and G. Plate 55, figs. 1-3 Fronds forming inconspicuous cushions, 0.75-1.5 mm. diam. ; pros- trate portion composed of very closely compact regularly radiating filaments; erect filaments densely crowded, cylindrical, all fructi- ferous; true hairs unknown; cells of creeping filaments 4/i diaijj., quadrate in the center of the frond, slightly longer at the margin ; ce Is of erect filaments the same in size as in the creeping filaments ; zoo- sporangia and "ascocysts" unknown; gametangia on 4-8-celled pedicels, cylindrical, densely crowded, 34-40/* long, 4-4. 5/t broad. Growing on the cysts of Egregia Menziesii. Cast ashore near the "Cliff House," San Francisco, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 336, pi. 34, figs. 1-3. Myrionema minutissimum may be distinguished by the absence of sterile erect filaments and hairs, and by the small sized, very densely crowded pedicellate gametangia. Thus far no "ascocysts" nor zoo- sporangia have been observed. In its phylogenetic relationship it would seem to stand equally close to M. corunnae and M. foecundum, differing from each, however, in details of measurements. The erect filaments are of the same diameter at their bases as the length of the cells in the creeping filaments. Since the lower part of each remains sterile, it gives the mass of cells the appearance of being parenchy- matous. 1925] Setchell-Gardner: Melanophyceae 465 Myrionema balticum (Reinke) Foslie Myrionema (Phycocelis) balticum (Reinke) Foslie. New or Crit. Norw. Alg., 1894, p. 17. Ascocyclus balticus Reinke, Algenfl. westl. Ostsee, 1889a, p. 45, Atlas, 1889, p. 19, pi. 16. figs. 1-4. A discussion of the reasons for rejecting Ascocyclus may be found under Myrionema foecundum. The chief character which Magnus used in establishing his new genus Ascocyclus, viz., the "ascocysts," docs not appear in the diagnosis given by Reinke for A. balticus. It seems to us clearly that Reinke 's plant is a Myrionema, and we are grouping two distinct forms under the species which, to us, seem amply different from the typical form, as described and figured by Reinke. 12. Myrionema balticum f. californicum S. and G. Fronds forming microscopic cushions, 0.5-0.75 mm. diam. ; pros- trate portion composed of relatively straight, regularly radiating fila- ments; erect filaments unbranched, attenuated slightly upwards and at the base, blunt, 75-125/a long; true hairs 0.5-1 mm. long; cells of prostrate filaments 6-7/i diam., 1-1.5 times as long; cells of erect filaments 7.5-10/* diam., mostly quadrate ; cells of true hairs 4-5. 5/* diam., 6-10 times as long above, 3-5 times as long at the base, quadrate a few cells above the base ; chromatophores irregularly broken bands ; zoosporangia unknown ; gametangia on 1-3-celled pedicels, either from the creeping filaments, or lateral at the base of the erect filaments, cylindrical, 90-130/* long, 7-9/* broad. Growing on the outer ends of the leaves of Phyllospadix sp. Point Carmel, Monterey County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 341. We have no authentic measurements of Ascocyclus balticus Reinke. He does not mention the size of the gametangia either in the Atlas (1889, pi. 16, figs. 1-4), where he first describes and figures it, or in the Algenflora (1889a, p. 45) where he describes it again. Foslie (1894, p. 17, repr.) made the combination Myrionema. {Phycocelis) balticum., but he gave no measurements for any of its parts. De-Toni (1895, p. 581) lists the plant as Phycocelis (?) baltica (Reinke) Foslie, but gives no additional information concerning the size of the game- tangia. Kylin (1907, p. 35) lists a plant under Myrionema balticum 466 University of California Publications in Botany [ V( >l. 8 (Rke.) Fosl. which he found growing on Delesseria sanguinea at 15 m. depth. He states that the assimilation filaments are 100-125/* long and -4-6/* broad, but gives no measurements for the gametangia. The plant which we have diagnosed is doubtless very closely related to Reinke's Ascocyclus balticus. Tt may be a distinct species, but we have listed it as a form of balticus until we can obtain more knowledge of its variation, and more authentic information concern- ing the type material of balticus. The particular distinguishing character is the very frequent occurrence of lateral gametangia arising very close to the base of the vegetative filaments. 13. Myrionema balticum f. pedicellatum S. and G. Fronds forming microscopic cushions 150-400/* (up to 800/x) diam., very irregular in outline ; erect filaments short-piliferous in part, 140-160/* long; true hairs unknown; zoosporangia (?) narrowly clavate, sessile on the creeping filaments, or terminal on the erect fila- ments ; gametangia terminal on 10-15-celled pedicels, rarely sessile, up to 150/* long. Growing on the outer ends of the leaves of Phyllospadix sp. Point Carmel, Monterey County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 341, pi. 33, figs. 11, 12. This form differs from the species in the absence of hairs, in having a few very short-piliferous, vegetative filaments, and in having the gametangia on long pedicels. In this form, as in many others, the correct interpretation of the zoosporangia mentioned above in the diagnosis is an open question. They are filled with dense cell contents and, although none has been observed to produce zoospores, they appear as though they might later. Whatever their nature may finally prove to be, they represent another character in which this form differs from the species. Reinke neither figures nor describes them in Ascocyclus balticus. They appear on the same individuals with the gametangia or on different individuals. 1925J Setchell-Gardner: Melanophyceae 467 14. Myrionema compsonematoides S. and G. Plate 55, fig. 10 Fronds forming microscopic cushions irregular in outline, 200-400/* diam., prostrate portion composed of distorted filaments; erect fila- ments cylindrical, unbranched, 70-80/t high ; hairs unknown ; cells of creeping filaments considerably irregular in shape and size, 4-6/i diam. ; cells of erect filaments cylindrical, 5-7/x diam., quadrate; "ascocysts" sparse, terminal or at times lateral near the base of the erect filaments; gametangia fusiform to clavate, sessile or short-pedicellate, 50-65/;. long, 7-10/i broad ; loeuli uniseriate below, biseriate above. Growing on the blade of Lamina ria complanata. Friday Harbor, Washington. Setchell and Gardner, Phyc. Cont. 1 1, 1922, p. 343, pi. 34, fig. 10. Myrionema compsonematoides seems to be a modification of M. foecwndum. There is a slight sterilization of tissue in that a few of the erect filaments remain sterile and develop slightly beyond the gametangia, and in that a part of the gametangia are pedicellate. It seems to tend toward Compsonema in these respects, and in having biseriate gametangia in part. 15. Myrionema hecatonematoides S. and G. Plate 55, fig. 11 Fronds forming thin light brown cushions, more or less circular in outline, up to 6 mm. diam. ; prostrate portion composed of distorted, irregularly radiating filaments, mostly monostromatic, in part distro- matic, producing a few penetrating peg-like rhizoids below; erect filaments unbranched, cylindrical, 240-260/t long ; true hairs absent ; cells of erect filaments 6.5-7. 5/* diam., 2-3 times as long; zoosporangia unknown ; gametangia sessile or on 1-2-celled pedicels on the creeping filaments, cylindrical to fusiform, 45-55/* long, 7-8. 5/t broad ; loeuli mostly uniseriate, biseriate in widest part of the gametangia. Growing on the pneumatocysts of Nereocystis Luctkeana. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 343, pi. 34, fig. 11. As indicated by the partially distromatic base, the slight tendency to biseriate loeuli in the gametangia, and the sterilization of a part of the erect filaments, this species of Myrionema seems to be differen- tiating in the direction of the genus Heeatonema. We are retaining 4:68 University of California Publications in Botany f VoL - s it in the genus Myrionema because of the preponderance in number of uniseriate loculi and monostromatic filaments in the prostrate portion. 16. Myrionema attenuatum S. and G. Plate 55, figs. 6, 7 Fronds forming microscopic, more or less confluent strata on the surface of the host; prostrate portion composed of regularly radiating filaments; erect filaments unbranched, attenuated gradually upwards and at the base, 180-220/* long ; true hairs absent ; cells of creeping filaments, 4.5-5.5/* diam., nearly quadrate ; cells of erect filaments 6-7/^ diam., variable in length; zoosporangia unknown; gametangia sparse, cylindrical to slightly fusiform, on shorter or longer pedicels, 30-45/x long, 4.5-6/1 broad. Growing on the stipes of Macrocystis pyrifera, Lammaria Farlowii, and on the sterile base of Gigartina radula f. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 344, pi. 34, figs. 6, 7. It seems that sterilization has proceeded in this species of Myrionema until about two-thirds of the erect filaments no longer produce gametangia, and these filaments are several times longer than the gametangia. The cells are slightly doliiform in some collections. 17. Myrionema attenuatum f. doliiforme S. and G. (Orthog. mut.) Plate 55, figs. 4, 5 Fronds forming small, irregular tufts among other epiphytic algae ; erect filaments up to 275/x long; cells of erect filaments quadrate to 3 times as long as broad, decidedly doliiform; gametangia occasion- ally small and terminal on the long erect filaments, otherwise as the species. Growing on the stipe of Macrocystis pyrifera. Carmel Bay, Monterey County, California. Myrionema attenuatum f . doliiformum Setchell and Gardner, Phyc. Cont. II, 1922, p. 344, pi. 34, figs. 4, 5. The chief distinction between this form and the species, M. attenuatum, is the decided doliiform character of the erect filaments. Small terminal gametangia appear very frequently on the long, erect filaments. 19-25] Setchell^Gardner: Melanophyceae 469 18. Myrionema phyllophilum S. and G. Fronds forming cushions more or less circular in outline, 400-800^ diam. ; creeping portion composed of straight filaments radiating from the center, closely crowded; erect filaments unbranched, cylindrical, constricted at the base, 110-130/x long, attenuated upwards, piliferous in part : true hairs scattered promiscuously among the erect filaments and gametangia; cells of creeping filaments cylindrical, 4-5/x diam., 2-4 times as long; cells of erect filaments 8-9^ diam., quadrate, to 2.5 times as long; cells of the hairs quadrate at the base, 5-6/* diam., up to 20 times as long in the upper part ; zoosporangia ( ?) broadly clavate, sessile or terminal on longer or shorter pedicels, 50-70/x long, 14-18/x broad at the outer end; gametangia cylindrical, mostly on 1-2-celled pedicels, on the creeping filaments, or more rarely lateral on the base of the erect filaments, blunt, 90-130/x long, 9-11/z broad; loculi uniseriate. Growing on the outer ends of the blades of eel-grass, in the lower littoral and upper sublittoral belts. Sitka, Alaska. Setchell and Gardner, Phyc. Cont. II, 1922, p. 344, pi. 32, figs. 1-5. In this species about one-third of the erect filaments remain sterile, and extend beyond the gametangia. The cells in the creeping filaments are relatively long. The filaments are closely crowded, but distinct and readily separable. All the erect filaments are abruptly narrowed at the base. On account of the long creeping cells, the erect filaments are very much less crowded than is usually the case in most Myrio- nemas. The erect filaments do not arise successively toward the margin of the nearly mature plants, cells here and there developing an erect filament to the length of several cells before the intervening cells show signs of farther growth. The gametangia are typical of the genus, with uniseriate loculi and with mostly horizontal, cross cell walls. The zoosporangia (?) are fairly abundant, and are either with the game- tangia on the same individual or on separate individuals. We ques- tion these structures as being functional since none of them has been observed to produce zoospores. They have the appearance of being abortive organs, either zoosporangia or gametangia, more likely the latter, and are probably like the organs observed by Magnus which induced him to establish the genus Ascoeyclus. 470 University of California Publications in Botany [Vol.8 19. Myrionema setiferum S. and G. Fronds forming small, dense, irregular tufts or cushions, 0.5-0.75 mm. diam. ; prostrate portion composed of regularly radiating, closely compacted filaments, 4-5/t diam., 1.5-3 times as long toward the margin ; true hairs unknown ; erect filaments numerous, unbranched, gradually attenuated toward the apex, long-pilif erous, 700-800/* long ; cells of erect filaments cylindrical, 4-5/x diam., 2-3 times as long ; zoosporangia broadly clavate, 80-90^ long, 18-24//. broad ; gametangia cylindrical, 150-200/x. long, 7-8.5^ broad, sessile, or on short pedicels on the erect filaments or rarely on the creeping filaments ; loculi uniseriate. Growing on the outer ends of the leaves of eel-grass, in company with other small algae, in the lower littoral belt. Sitka, Alaska. Setchell and Gardner, Phyc. Cont. II, 1922, p. 345, pi. 32, fig. 8. Myrionema setiferum is undoubtedly very closely related to M. phyllophilum on the one hand, and to M. foecundum f. calif ornicum on the other. From the former, it differs in having longer and nar- rower erect filaments with longer cells and in having much longer and narrower gametangia which are more frequently borne laterally on the erect filaments, and often fascieulately branched from a short pedicel. From the latter, it differs in having functional zoosporangia, one of the few instances in which we have actually seen the undoubted zoospores, in having longer and setiferous erect filaments, and longer and slightly narrower gametangia. Myrionema globosum (Reinke) Foslie Myrionema (Phycoeelis) globosum Foslie, New or Crit. Norw. Algae, 1894, p. 17. Ascocyclus globosus Reinke, Algenfl. westl. Ostsee, 1889a, p. 46, Atlas, 1889, p. 20, pi. 17. Microspongium globosum Reinke, Die brauenen Alg., 1888, p. 20. Typical M. globosum, as we understand this plant, has not been reported from our territory, but we have described a forma affine which seems very closely related to it. 1925] SetcheU-Qardncr: MeUinoplnjcrar 471 20. Myrionema globosum f. affine S. and G. Fronds microscopic 200-300//. diam., irregular in outline, inter- mixed with other small algae; prostrate filaments much distorted, freely branched; erect filaments fasciculately branched, 100-140/t long, tapering upwards, piliferous in part ; true hairs 300-400/1 long ; cells of prostrate filaments 4-6/t diam., variable in shape ; cells of erect filaments cylindrical, 4-5/x diam. ; zoosporangia unknown ; game- tangia cylindrical, 35-45/*. long, 4-4.5/x broad, sessile and lateral on the erect filaments or rarely terminal ; loculi uniseriate. Growing on the leaves of eel-grass. Sitka, Alaska, to central Cali- fornia (Point Carmel, Monterey County). Setchell and Gardner, Phyc. Cont. II, 1922, p. 342, pi. 33, fig. 16. We have no measurements in detail for the parts of Myrionema globosum, this seemingly distinct species. Reinke (1889a, p. 46) states that the plants are one-half to one millimeter in diameter. Our plants seem much smaller, in general. It is difficult to make out the size on account of their being so intimately intermixed with other small algae. This is the case in both collections of plants which we have examined, one from Sitka, Alaska, and the other from Point Carmel, California. 21. Myrionema strangulans Grev. (in scnsit strict lore) Plate 35, fig. 12, and plate 40, fig. 51 Fronds forming very irregularly shaped cushions, 1-2 mm. diam., firmly adhering to the surface of the host, without rhizoids ; prostrate portion composed of regularly radiating filaments with apical growth and branching, producing dichotomous filaments, strictly monostro- matic; cells of prostrate filaments 4-4.5/x diam., 1-2 times as long, cylindrical or slightly doliiform; hairs, 1-1.5 mm. long, cells of hairs up to 100/A long at the apices, 5/* diam., quadrate and with short sheath at the base ; erect filaments clavate, longest in the center of the frond, diminishing in length toward the periphery, 55-65/* (up to 110/x) long, 4/* diam. at the bases, Bp. at the apices; basal cells cylindrical, terminal cells subspherical ; chromatophores spherical, several in a cell; unilocular zoosporangia numerous, interspersed among the erect filaments, sessile or on 1-celled pedicels on the creeping filaments, ellipsoidal to obovoid, 38-46/* (up to 60/*) long. 18-24/* broad; game- tangia unknown. 472 University of California Publications in Botany [Vol. 8 Growing- on the base of Viva sp. in the lower littoral belt. Sitka, Alaska, to central California (Carmel Bay). Greville, Scot. Crypt. Flora, vol. 5, 1827, pi. 300; Harvey, Phyc. Brit., vol. 3, 1851, pi. 280; Saunders, Alg. Harriman Exp., 1901, p. 423, pi. 46, figs. 1, 2; Setchell and Gardner, Alg. N.W. Amer, 1903, p. 249; Tilden, Amer. Alg. (Exsicc.), no. 350. Not Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1795 (a mixture of several Myrionemataceae) . Myrionema strangulans Greville has been variously interpreted and the name has frequently been placed as a synonym under M. vulgare Thnret. In the attempt to determine the exact nature of M. strangulans, specimens from Greville 's herbarium, from the western isles of Scotland (but not from Appin, the technical type locality) have been obtained (through the kindness of Professor Isaac Bayley Balfour) and carefully studied. From these the description given above has been drawn up. We find specimens on Viva which agree with these Greville specimens in every detail, except that the zoo- sporangia in our plants average slightly larger. 22. Myrionema obscurum S. and G. Fronds microscopic, distributed more or less irregularly on the surface of the host ; creeping filaments much branched, very compact and contorted, forming a pseudoparenchymatous layer ; erect filaments sparse, unbranched, 40-70/x high ; true hairs unknown ; cells of creep- ing filaments extremely irregular in shape and size, 4.5-10/x diam. ; cells of erect filaments cylindrical, 6.5-7.5/a diam., quadrate or nearly so ; zoosporangia very numerous, sessile on the creeping filaments, ovoid to broadly clavate, 45-65/x long, 18-30^ broad ; gametangia unknown. Growing on the blade of Costaria costata. Moss Beach, San Mateo County, California. Setchell and Gardner, Phyc. Cont. II, 1922, p. 346, pi. 32, fig. 6. Myrionema obscurum seems decidedly distinct from all other species of the genus. It seems to be a near relative of M. strangulans Grev. The basal layer is very compact, and adheres very firmly to the host, following closely the irregularities on its surface. We are unable to make out whether or not the epiphyte causes the death of the cells of the host, which seems- to be quite frequent in all of the plants we have observed. The creeping filaments push among the dead cells of the host at times. The zoosporangia are numerous and all sessile. 1925] Setchell-Gwrdner: Melanophyceae 473 7. Compsonema Kuek. Fronds forming small cushions more or less circular in outline. composed of a prostrate creeping, primary portion, giving rise to numerous erect unbranched or branched filaments, with or without hairs; prostrate portion monostromatic, composed of filaments quite regularly radiating from a common center and usually with sub- terminal branching, closely crowded ; reproduction by unilocular zoosporangia and plurilocular gametangia with mostly pluriseriate loculi. Kuckuck, Beitriige zur Kennt. Meeresalg., 1899, pp. 56-58 (90-92 inrepr.),pl. VI (12), figs. 6-9. The type species of Compsonema is C. gracile and the type locality is Rovigno, on the east coast of the Adriatic Sea, where it was found growing on stones in water 1-2 meters deep. The genus Compsonema was established by Kuckuck. We have accepted the genus to include forms which, like Myrioncma, start from a single cell that, by divisions, soon originates a series of closely crowded, radiating filaments, forming a monostromatic basal layer more or less circular in outline; and which, unlike Myrionema, pro- duce on erect filaments numerous gametangia that develop pluriseriate loculi. The genus is very closely related to the genus Myrionema on one side and to Hecatom ma on the other. Key to the Species 1. Erect sterile filaments branched 2 1. Erect sterile filaments simple 5 2. Creeping filaments with few subulate rhizoids 7. C. fasciculatum (p. 478) 2. Creeping filaments without rhizoids 3 3. Sterile branches short, high up, secund 4 ■3. Sterile branches long, low down, not secund 13. C. intricatum (p. 482) 4. Erect filaments piliferous in part 17. C. ramulosum (p. 484) 4. Erect filaments not piliferous 12. C. sporangiiferum (p. 481) 5. Gametangia all terminal 6 5. Gametangia terminal and lateral or intercalary 14 6. Gametangia all sessile 7 6. Gametangia pedicellate or in part sessile 8 7. All erect filaments completely transformed into gametangia. ...2. C. sessile (p. 474) 7. Gametangia intermixed with sterile filaments 6. C. tenue (p. 478) 8. Prostrate filaments producing penetrating rhizoids 9 8. Prostrate filaments not producing rhizoids 10 9. Rhizoids abundant, straight, penetrating deeply into the host 11. C. streblonematoides (p. 481) 9. Rhizoids contorted, not penetrating deeply 5. C. pusillum (p. 477) 10. Gametangia very abundant, of uniform height .4. C. fructuosum (p. 476) 10. Gametangia with pedicels very variable in height 11 474 University of California Publications in Botany [Vol. 8 11. Gametangia cylindrical or very slightly tapering 12 11. Gametangia decidedly fusiform 13 12. Gametangia mostly on 1-2-celled pedicels 1. C. myrionematoides (p. 474) 12. Pedicels of gametangia very variable in height .3. C. nummuloides (p. 475) 13. Gametangia 24-30/u long, 6-8ju broad 18. C. immixtum (p. 487) 13. Gametangia 45-65^ long, 15-1Sm broad 9. C. coniferum fp. 479) 13. Gametangia 50-75m long, 8-11/x broad 8. C. dubium (p. 479) 13. Gametangia 60-130^ long, 18-28/* broad 10. C. serpens (p. 480) 14. Erect filaments in part piliferous 16. C. speciosum f. piliferum (p. 485) 14. Erect filaments not piliferous 15 15. Terminal gametangia up to 400^ long 14. C. secundum fp. 483) 15. Terminal gametangia up to 110/x long 15. C. secundum f. terminale (p. 484) 1. Compsonema myrionematoides S. and G. Fronds forming small, circular to irregular cushions, 1-2.5 mm. diam., prostrate portion strictly monostromatic, composed of closely crowded, contorted, radiating filaments ; erect filaments unbranched, cylindrical, 110-130/*. long; true hairs absent; cells of creeping fila- ments cylindrical to irregular in form ; cells of erect filaments cylin- drical, 7.5-8.5//, diam., 1-2 times as long, terminal often several times as long ; chromatophore a slightly broken band in each cell ; zoo- sporangia (?) broadly clavate, sessile or on short pedicels on the creeping filaments, 35-50/x. long, 20-28/*, broad ; gametangia numerous, cylindrical to fusiform, sessile or on short pedicels on the creeping filaments, 50-7 0/x (up to 100/*) long, 9-12/*. broad. Growing on the stipe of Nereocystis Luetkeana. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 361, pi. 36. fig. 1. There is but little choice as to the generic position of this small alga. It forms small brown tufts or cushions of loose filaments. Very frequently the gametangia are wider in the middle than at the ends and have decidedly biseriate loculi. Because of the more or less loose tufted character of the erect filaments and the partially biseriate gametangia we have placed it with the genus Compsonema. 2. Compsonema sessile S. and G. Fronds forming inconspicuous cushions up to 1.5 mm. diam., cir- cular to irregular in outline ; creeping filaments contorted, closely crowded, adhering firmly to the host, without rhizoids ; erect filaments sparse, 20-24/*. high; true hairs absent; cells of creeping filaments irregular in shape and size, 9-12/* diam. ; chromatophores band-shaped, one in each cell ; zoosporangia unknown ; gametangia numerous, sessile 1925] SetcheU-Gardner: Melanophyceae 475 on the creeping filaments, cone-shaped to fusiform, 16-22/x long, 11-14/* broad, many walls of the loculi oblique. Growing on the lamina of Hedopln/llum sessttt near the outer end. Neah Bay, Washington. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 358, pi. 39, fig. 6. Compsonema sessile represents the extreme of the genus in the direction of simplicity. There are no hair filaments, and it is doubtful whether any of the erect filaments remain steiile. At times even the cells of the basal layer seem to become a part of the gametangia, at least more than one cell in the basal layer may be at the base of a single gametangium. There would seem to be only a single modifica- tion of Myrionema of the simplest type to produce this species, viz., the change in the type of gametangium from uniseriate to pluri- seriate. If we hold to the principal distinction between Compsonema and Myrionema, it will be necessary to ally this species with the former. It was found growing in company with StrebJonema aeci- d hides f. pacifieum S. and G., which gave to the host a roughened appearance, otherwise it probably would not be noticeable without microscopic examination. 3. Compsonema nummuloides S. and G. Fronds forming thin circular cushions, 7-12 mm. diam. ; prostrate portion composed of radiating, much crisped, closely crowded filaments following closely the irregularities of the host; erect filaments unbranched, numerous, 300-400^ long, tapering slightly at both ends; true hairs absent ; cells of creeping filaments very irregular in shape and size ; cells of erect filaments cylindrical, (i-7/x diam. at the base, 3-8 times as long, 9-llfi diam. in the widest part, quadrate apical cell 6-7/x diam., 3-8 times as long; zoosporangia unknown; gametangia mostly sessile or on short pedicels, some terminal on the erect filaments, 70-90ju (up to 140/*) long, 11-lSju broad ; loculi mostly 2-seriate. Growing on the pneumatocysts of Nereoeystk Luetkeana. Moss Beach, San Mateo County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 359, pi. 35, figs. 5, 6. Compsonema nummidoides forms thin, light brown cushions on the pneumatocysts of the host. The whole cushion is a single plant, circular in outline and attaining a diameter of twelve or more milli- meters, although the majority of the plants are much smaller. Like a typical Myrionema the plant starts from a single cell and by divi- sions forms a series of filaments radiating in all directions with apical 476 r nirtrsih/ of California Publications in Botanij L VoL - 8 growth and dichotomous branching, by the splitting of the apical cell, thus forming a continuous monostromatic layer of cells. Beginning in the center, practically every cell in each radiating filament of the basal layer gives rise successively to an erect filament, about two-thirds of which are fructiferous, the others remaining sterile. A very large majority of the gametangia are sessile or on short pedicels, thus occupying a zone near the creeping filaments, a myrionematoid char- acter, but the remainder of the filaments continue to grow and attain a length of 300//. to 400/x. Some of these filaments are terminated by relatively short gametangia. This is one of several species with this general method of development which has been found growing only on the above mentioned host, differing from each other and from the type of the genus, C. gracile Kuckuck, in the size of the plant as a whole, in details of dimensions of their parts, in the presence or absence of zoosporangia and hairs, in the position of the gametangia, and in the relative amount of sterile filaments. It seems to form a fairly compact group but with overlappings, however, in the genera Stre- blonema, Myrionema, and Hecatoncma. 4. Compsonema fructuosum S. and G. Fronds forming circular or somewhat irregular cushions, 4-5 mm. diam. ; prostrate portion composed of tortuous creeping filaments fol- lowing closely the irregularities of the surface of the host ; erect fila- ments unbranched, densely crowded, 190-230/*. long, cells cylindrical, 8-9/x diam., 2-4 times as long below, 1-2 times above; chromatophores band-shaped ; hairs absent ; zoosporangia sparse, broadly clavate, terminal on short pedicels from the creeping filaments, 55-65/x (up to lOOju) long, 22-28/x broad; gametangia terminal on erect filaments. cylindrical to slightly fusiform, blunt, 80-120/;. long, 12-1 6p. broad; loculi 2-4-seriate. Growing on the pneumatocysts of Nereoeystis Luetkeana. Tomales Bay, Marin County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 355, pi. 36, fig. 2. Compsonema fructuosum departs from the type of the genus, C. gracile Kuckuck (1899, p. 90, pi. VI (12), figs. 6-9), in having a differ- ent host, no hairs, fewer and shorter sterile filaments, more gametangia, and in having differences in details of measurements. In the abun- dance and position of the fruit and in the definite circular thallus with compact monostromatic basal layer of filaments, it approaches 1925] SetcheU-Gardner: Mclanophyrecu 477 very closely to a typical Myrionema, but the multiseriate gametangia represent a stage in developmenl which we have not admitted into that genus. The gametangia occupy a definite zone on the outer ends of the erect filaments. Practically all of the erect filaments bear gametangia. At times the basal layer seems to be distromatic, or even poly- stromatic, but careful investigation shows that this appearance is occasioned by the overlapping creeping filaments whieh follow the v< ry irregular surface of the host, due to the death of its surface layer of cells, probably before the epiphyte took possession. The zoosporangia are very sparse, but specimens have been seen which show undoubted zoospores. 5. Compsonema pusillum 8. and G. Plate 54, fig. 3 Fronds forming thin, light brown cushions, irregular in outline, 3-4 mm. diam. ; prostrate portion composed of very tortuous, branched filaments following closely the irregularities of the host, producing a few filaments from the under side penetrating among the cells of the host ; erect filaments simple, cylindrical, 270-300/t long ; cells of creep- ing filaments very irregular in shape and size ; cells of erect filaments cylindrical, 6-7/* diam., 3-5 times as long; chromatophore a slightly broken band ; true hairs absent ; zoosporangia spherical to ovoid, sessile or on short pedicels on the creeping filaments, 18-25/j. (up to 40/0 long, 16-22//. broad; gametangia cylindrico-conical to blunt fusiform, sessile or on short pedicels on the creeping filaments, 40-60/* long, 10-12/* broad ; loculi 1-2-seriate. Growing on the pneumatocysts of Ncrcocystis Luetkeana. Carmel l Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 356, pi. 37, fig. 3. Compsonema pusillum seems to overlap each of several genera. Its zoosporangia, gametangia, and penetrating filaments ally it with Streblonema. Sauvageau (1897) has figured a Myrionema with rhizoids. The gametangia are very small and have about as many uniseriate as biseriate loculi. It might be considered a very minute creeping Ectocarpus. We have placed it with the genus Compsonema on account of the scarcity of fruit in proportion to the sterile tissue, and on account of the monostromatic base, believing that the penetrat- ing filaments are more or less accidental on account of the injured condition of the surface layer of cells of the host. 478 University of California Publications in Botany L VoL - 8 6. Compsonema tenue S. and G. Plate 54, fig. 6 Fronds microscopic, forming a more or less continuous stratum on the host among other small algae; creeping filaments numerous, branched, following closely the irregularities of the surface of the host; erect filaments sparse, unbranched, tapering slightly at both ends, 80-100/t high ; hair filaments sparse ; cells of creeping filaments irregular in shape and size; cells of erect filaments cylindrical, not constricted, 7-8/* cliam., 1.5-2.5 times as long; cells of hair filaments cylindrical, 5.5-6.5/1. diam., quadrate or less below, 7-10 times as long above; zoosporangia (?) clavate, 24-28/x long, 8-10/x, broad, sparse; gametangia fusiform to irregularly conical, sessile on the creeping filaments, numerous, 22-28/x, long, 7-11/t broad ; loculi mostly 2-seriate. Growing on the fruiting fronds of Cystophyllum geminatum. Sitka, Alaska. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 359, pi. 37, fig. 6. Compsonema tenue, although scarcely typical of the genus, seems to possess affinities as closely allied to Compsonema as to any other genus, and hence we have placed it here for the present. 7. Compsonema fasciculatum S. and G. Fronds forming microscopic tufts more or less confluent ; creeping filaments tortuous, adhering to the host without rhizoicls; erect fila- ments branched near the base, 90-130/* long, tapering at the apex and at the base, not piliferous, constricted at the cross-walls; true hairs absent ; cells of erect filaments 10-14/x diam. in the widest part, 1-2.5 times as long ; zoosporangia unknown ; gametangia somewhat irregular in form, conical to fusiform, 35-45/x long, 12-18/t broad, mostly lateral on short pedicels. Growing on Gigartina radula f. Pacific Grove, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 360, pi. 38, figs. 7-9. This species is allied here rather than with Ectocarpu-s because of its extremely small size and its creeping fasciculate habit. 1925] Setchell-Gardner: Melanophyceae 479 8. Compsonema dubium S. and G. Fronds diminutive, forming' small tufts among other microscopic algae on the host ; prostrate portion composed of much contorted, branched, creeping filaments; erect filaments unbranched, almost cylindrical throughout, 275-350^ long, not piliferous; true hairs absent; cells of creeping filaments very irregular in form and size; cells of erect filaments cylindrical, 7-9/* diam., slightly less at the base and the apex of the filament, 2-3 times as long; zoosporangia (?) ellipsoidal, terminal on short pedicels, 16-22/x long, 12-15/* broad; gametangia cylindrical to slightly fusiform, terminal on longer or shorter pedicels from the creeping filaments, rarely terminal on the long erect filaments, 50-75/t long, 8-11/* broad ; loculi mostly 2-seriate. Growing on the sterile base of Gigari ma radula f. Cypress Point, Monterey County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 365, pi. 38, fig. 6. Compsonema dubium, like C. coniferum and C. serpens, has close affinities with small members of the genus Ectocarpus. Its relatively long and narrow gametangia distinguish it from the two species here mentioned with which it is frequently associated. 9. Compsonema coniferum S. and G. Fronds microscopic, more or less confluent, forming a velvety stratum on the host among other minute algae; prostrate portion pro- fusely branched, much contorted; erect filaments numerous, simple, slightly attenuated at both ends, not piliferous, 150-200/* high; true hairs absent ; cells of creeping filaments irregular in shape and size, 6-7. 5ti diam., 9-12/* long; cells of erect filaments cylindrical to slightly doliiform, 8-10/* diam. in widest part, 1-1.5 times as long; chromato- phores band-shaped; zoosporangia unknown; gametangia sessile or on short pedicels on the creeping filaments, narrowly conical, 45-65/* long, 15-18/* broad. Growing on the sterile base of Giyartina radula f. Cypress Point, Monterey County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 365, pi. 38, fig. 3. Compsonema coniferum seems to be a very close relative of C. fasciculatum, found growing on the same host. The fronds are more extensive, unbranched, and have larger gametangia. We consider this species of Compsonema a very near approach to an Eetocarpus of 480 University of California Publications in Botamy 1 Vol. 8 extremely small size. We are inclined toward the genus Compsonema as a more appropriate genus for this species than Eetocarpus on account of its extensive creeping habit and prolific production from the creeping filaments of nearly sessile gametangia and many sterile erect filaments. It is to be found quite commonly associated on the same host with several other species of Myrionemataceae. 10. Compsonema serpens S. and G. Fronds forming a more or less continuous expanded stratum on the surface of the host ; creeping portion composed of much contorted, irregularly branched filaments; erect filaments unbranched, or rarely with a few subulate branches above, tapering slightly at the base, gradually attenuated in the upper part, not piliferous, 375-425/* long ; true hairs absent; cells of creeping filaments cylindrical to irregular, 10-13/* diam., 0.5-3 times as long; cells of erect filaments cylindrical, not constricted at the dissepiments, 5.5-8. 5/* diam. at the base, 10-17/t diam. in the widest part, 1.25-2.5 times as long ; chromatophores band- shaped, densely crowded in the young cells, more numerous and separated in the old cells; zoosporangia unknown; gametangia cylin- drico-conical, terminal on shorter or longer pedicels from the creeping filaments, rarely terminal on the long filaments, 60-130/* long, 18-28/* broad. Growing on the sterile base of Gigartina raduia f ., in company with several other small algae. Cypress Point, Monterey County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 363, pi. 39, fig. 7. This diminutive member of the Melanophyceae is on the border line between Compsonema and Eetocarpus. Our comprehension of these two genera, so far as the vegetative portion is concerned, is that a typical Compsonema, starting from a single cell, develops an exten- sive, creeping, attaching mass of filaments, radiating in all directions from the origin, which later produces very numerous erect filaments, some of which remain sterile and others become fructiferous; and that a typical Eetocarpus, starting likewise from a single cell, develops a rather insignificant series of attaching filaments, more or less rhizoidal in nature, and a relatively extensive system of erect filaments finally producing the fruit. Starting with this conception concerning the two genera, we are assuming that differentiation has proceeded from Compsonema in the direction of the reduction of the creeping portion, to a more extensive development of the erect portion, finally resulting in an Eetocarpus. 1925] Setchell-Gardner: Melanophyceae 481 The species under consideration has a relatively extensive system of creeping filaments but very much less so than that of a typical Comp- sonema. The creeping filaments do not form a solid disk, but those from different plants are so intertwined that it is impossible to deter- mine the limits of a distinct individual. In this character it resembles an Ectocarpas. The gametangia are like those of a typical Ectocarpas, but since they are mostly short-pedicellate and spring directly from the creeping filaments, as is the case in a typical Compsonema, and because of the relatively extensive system of creeping filaments, we have placed it in the latter genus. 11. Compsonema streblonematoides S. and G. Fronds forming thin, circular cushions, 3-5 mm. diam. ; prostrate portion composed of filaments following closely the irregularities of the host, giving rise below to numerous rhizoidal filaments penetrating to a depth of 300-350/*. among the cells of the host, and above to sterile filaments and gametangia; erect filaments 140-160/* long; cells of erect filaments cylindrical. 8.5-10/* diam., 1.5-3 times as long ; zoo- sporangia unknown; gametangia mostly cylindrical, short-pedicellate, 55-70/x long, ll-14/i broad ; loculi mostly 2-seriate. Growing on the pneumatocysts of Nereocystis Luetkeana. Mouth of Tomales Bay, Marin County, California. • Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 353, pi. 35, fig. 4. Compsonema streblonematoides, like a Streoloncma, has a large part of the vegetative portion of the frond within the host -plant, the penetrating filaments extending into the host two or three times as far as the length of the external vegetative filaments. The whole frond, however, is the result of the growth of a single cell and forms a com- pact monostromatic basal disk. In the character of the gametangia it is close to a Myrionema, but in a very large majority of cases the loculi are biseriate. 12. Compsonema sporangiiferum S. and G. Fronds forming circular cushions, 1-4 mm. in diameter; prostrate portion composed of compact, irregular, more or less contorted fila- ments following the irregularities of the surface closely, monostro- matic or in part distromatic ; erect filaments 300-700/x long, un- branched or bearing above a few short, mostly secund, fructiferous ramuli ; cells of the erect filaments 6-8/*. diam. at the base, and 2-3 482 University of California Publications in Botany [Vol.8 times as long, 11-13/a diam. at the apex, and 1-1.5 times as long, many cells in the upper part of the filaments dividing 1-2 times longitudin- ally ; ehromatophores band-shaped, few in a cell ; zoosporangia very abundant, very variable in shape and size, narrowly ellipsoidal, clavate to broadly ovoid, 40-130/x long, 20-35/* wide, sessile or on short pedicels from the prostrate filaments, or lateral or terminal on the erect fila- ments ; gametangia rare, cylindrical, sessile or mostly short-pedicellate on the creeping filaments, rarely terminal and tuberculate on the erect filaments, 80-1 30/a long, 11-15/a broad. Growing on the pneumatocyst of Nere&cystis Luetkeana, Neah Bay, Washington. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 357, pi. 36, figs. 3-8. The unusual condition in Compsonema prevails in C. sporangi- iferum. The gametangia seem to be of rare occurrence. Examination of specimens of different sizes from different parts of the host revealed the presence of a large number of zoosporangia in all stages of develop- ment, many being empty. The gametangia are prevailingly near the base, although occasionally a complex one appears terminal on an erect filament. A very common character is the division of the cells in the upper parts of the filaments into 2-4 parts by longitudinal walls, producing a slight clavate condition in such filaments. We do not know, at present, the morphological significance of this condition. The zoosporangia and gametangia, so far as observed, develop on distinct individuals. 13. Compsonema intricatum S. and G. . Fronds 1.5-2.5 mm. high, more or less confluent, forming a con- tinuous velvety stratum of indefinite expansion; creeping filaments numerous, profusely branched, crooked; erect filaments numerous, unbranched or with a few short branches near the base, straight, grad- ually attenuated at the apices, piliferous in part ; true hairs absent ; cells of creeping filaments irregular in shape and size ; cells of erect filaments cylindrical, 8-9/* diam., 1-2.5 times as long, slightly narrower at the base of the filaments; ehromatophores band-shaped; zoo- sporangia ovoid to ellipsoidal, terminal on short filaments from the creeping filaments or lateral and sessile or pedicellate on the erect filaments near the base, 25-33/x long, 18-22/a broad ; gametangia ter- minal on short erect filaments from the creeping filaments, lateral and sessile or pedicellate near the base of the erect filaments or intercalary in the main erect filaments, 80-12G> (up to 175/*) long, 10-14^ broad; terminal and intercalary ones up to 60G> long ; loculi mostly 2-seriate. 1925] Setchell-Gardner: Melanophyceae 483 Growing on Fhcux furcatus f. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 354, pi. 35, figs. 1-3. Compsonema intricatum has diverged very far from a typical Compsonema. The size of a single individual at maturity cannot be ascertained, since 1 the creeping filaments, and to some extent the erect filaments, are thoroughly and inextricably intertwined, forming con- tinuous velvety strata on the surface of the host for several inches in expanse. The gametangia have, to a considerable degree, moved to the lateral position, the majority, however, remain terminal on long pedicels from the basal filaments. The lateral position is characteris- tic of Ectocarpus. The intercalary gametangia are typical of the genus Pylaiella. The creeping filaments are very numerous, although they do not radiate regularly side by side, thus forming a regular circular disk typical of the genus Myrionema. The zoosporangia (?) are numerous and develop with the gametangia on the same individual. It is exceedingly doubtful whether or not these structures are func- tional. At the stage in which the material was found, they were prac- tically all of the same size, and some of the accompanying gametangia were empty. There is no indication of the formation of zoospores. They are filled with what seem to be densely crowded, angular chro- matophores. The apical end is slightly beaked, and the outer wall dissolved, except a thin inner membrane. Their position on the same plants with gametangia, the peculiar open beak, and the fact that they all seem to be of the same age lead us to suspect that they are some sort of abnormalities, or pathological conditions of the gametangia, which will require much more investigation to interpret. We have retained this in the genus Compsonema on account of the prevailing simplicity of the erect filaments and the abundance of creeping fila- ments, practically every cell of which gives rise to an erect filament. 14. Compsonema secundum S. and G. Plate 54, figs. 1, 2 Fronds very small, usually more or less confluent ; prostrate por- tion monostromatic, composed of much crisped, branched filaments ; cells of creeping filaments irregular in shape, 7-8/*, diam., variable in length; erect filaments up to 1.25 mm. long, unbranched, or rarely with a few short fructiferous ramuli ; cells of erect filaments cylindrical, ~k~)-6.7)/x diam. at the base, 4.5-6 times as long, 5). 5-10. 5^ diam., in 484 University of California Publications in Botany [Vol, 8 widest part, terminal cells up to 9 times as long ; chromatophores band- shaped ; zoosporangia broadly clavate, sessile or on short pedicels from the creeping filaments, 60-90/* long, 22-28/*, broad; gametangia very variable, rarely on short pedicels on the creeping filaments, at times formed by transforming the upper portion of a long erect filament, at other times by developing lateral secund projections from part or all of the groups "of loculi in the long, terminal gametangia, up to 400ju, long, 11-15/* broad, blunt, clavate on the ends of the long filaments. Growing on the pneumatocyst of Nereocystis Luetkecmd. West coast of Washington (Moclips) to central California (Carmel Bay). Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 361, pi. 37, figs. 1, 2. Like Compsonema ramulosum, C. secundum has gametangia occupy- ing a variety of positions and assuming a diversity of forms. The form of gametangium which marks the species is the long terminal form which develops numerous, short, blunt-conical, seriate, secund, sessile gametangia, thus forming a complex composed of these game- tangia and the metamorphosed cells in the upper part of the filament. The end of the filament becomes reflexed, at times almost scorpioid. The specimens from Moclips, Washington, were taken in May and the gametangia were practically all empty. The other parts of the plants manifested signs of old age conditions. They seem to belong to this species, but the filaments have a considerably greater diameter. Typical zoosporangia with well-formed zoospores have been observed in this collection. 15. Compsonema secundum f. terminale S. and G. Plate 54, figs. 4, 5 Fronds forming a continuous velvety stratum of indefinite expanse, attached by very much distorted, branched, creeping filaments with numerous, penetrating, subulate rhizoids from the under surface; erect filaments 700-800^ (up to 1 mm.) long, unbranched, or produc- ing a few short secund fructiferous ramuli, straight, slightly tapering toward the base, cylindrical above ; cells of the creeping and rhizoidal portions irregular, variable in size; cells of the erect filaments cylindrical, not constricted, 6-8/t diam. at the base of the filaments, 2-3.5 times as long, 9-12/1 diam. above, 0.5-2 times as long; chromato- phores irregular bands, few in a cell ; zoosporangia ( ?) broadly clavate, sessile or on short pedicels from the creeping filaments, 50-60/t long, 1925 1 Setchell-Gardner: Melanophyceae 4sr> 18-22/*. broad, or terminal and seriate with lateral secund prolonga- tions; gametangia sessile near the outer ends of the erect filaments, or on short pedicels from the creeping filaments, 90-110/4. long, 15-18/* broad, nearly cylindrical, or terminal on the erect filaments, clavate, often with lateral, tuberculate, secund prolongations, variable in size. Growing on the pneumatocysts of Nereocystis Luetkeana. Central California (San Francisco and Pacific Grove). Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 366, pi. 37, figs. 4, 5. This form of Compsonema is to be distinguished by the great predominance of relatively short terminal gametangia, and many relatively large, lateral, secund, sessile gametangia which are curved upwards. There are a few sessile or short pedicellate examples at the base and a few lateral forms on short ramuli. Specimens taken in December had the larger number of gametangia empty. 16. Compsonema speciosum f. piliferum S. and G. Fronds forming cushions, more or less circular in outline, 5-8 mm. diam. ; creeping filaments very irregularly and profusely branched, following closely the irregularities on the surface of the host, at times short branches penetrating among the injured cells of the host ; erect filaments numerous, unbranched or bearing short, secund, fructiferous branches or an occasional piliferous branch, 500-800/*. long, tapering at the base, piliferous above ; cells of creeping filaments very irregular in shape and size; cells of erect filaments cylindrical, 6-7. 5/x diam. at the base, 9-12/*. diam. in the widest part, 1-3 times as long; cells of hairs 5-6. 5/x diam., up to 15 times as long; chromatophores band- shaped, few in a cell ; zoosporangia obovoid, 55-65/* long, 28-32/t broad, rare; gametangia sessile or on shorter or longer pedicels from the creeping filaments or lateral, secund, sessile or pedicellate on the erect filaments or, more rarely, intercalary ; terminal gametangia 60-100/* long, 14-18/* broad, secund gametangia variable, much smaller. Growing on the pneumatocysts of Nereocystis Luetkeana. Moclips, Washington. Setchell and Gardner, Phyc. Cont. Ill, 1922, p. 356, pi. 38, figs. 1, 2. Compsonema speciosum f. piliferum seems undoubtedly to be very closely related to Compsonema speciosum (Borg.) S. and G. comb, nov., (Myrionenm speciosum Borgesen, 1902, pp. 421-424, which is the Hecatonema speciosum (Borgesen) Cotton, 1912, p. 121 and the 486 University of California Publications in Botaiui [Vol.8 Hecatonema diffusum Kylin cf. Cotton, Inc. cit.), at least the two sets of plants agree very closely in general habit and structure as described by Bcirgesen. Judging alone from the description and figures the chief differences are that our plant has short penetrating rhizoids, many terminal and no lateral hairs, and larger gametangia. We are unable to say that the plant would produce rhizoids on an uninjured host. The surface cells of the host are destroyed under the epiphyte, and it is an open question whether the penetrating rhizoids are or are not the cause of their death. Borgesen does not give the dimensions of the whole plant, but he states that the plants form short dense mats on the conceptacles of Himanthalia lore a, which would indicate that they were quite small. Our plants form circular cushions or mats 5-8 mm. in diameter. Typical zoosporangia were found at the base of the erect filaments. The lateral secund gametangia, stand- ing nearly perpendicular to the filaments, the cells of which become the basal part of the gametangia, is a character not found in Comp- sonema gracile Kuckuck. 17. Compsonema ramulosum S. and G. Fronds forming circular cushions 5-7 mm. diam. ; prostrate por- tion monostromatic, composed of more or less contorted and irregularly branched radiating filaments; erect filaments piliferous in part, unbranched or bearing above short, mostly fructiferous and secund branches and occasional hairs, tapering slightly at the base ; cells of creeping filaments 6-7/* diam., quadrate or slightly longer than the diameter ; cells of erect filaments 6-7. 5/x, diam. at the base, slightly broader above, 1-3.5 times as long ; cells of the branches and hairs 4-6/* diam. ; chromatophores band-shaped, more or less broken and angular; zoosporangia (?) broadly clavate, terminal on longer or shorter pedicels, lateral and secund or whorled near the ends of the erect filaments; gametangia exceedingly variable in shape, size and position, sessile or pedicellate on the creeping filaments or lateral on the erect filaments or intercalary or terminal either on short lateral ramuli or on the main filaments, clavate, cylindrical to cylindrico- conical, up to 140/*, long, 16/* broad, loculi mostly 2-seriate. Growing on the pneumatocysts of Nereocystis Luetkeana. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont, III, 1922, p. 362, pi. 39, figs. 1-5. There exists in this species extreme variation in the form, size, and position of the gametangia. Some are sessile on the creeping filaments, 1925] Setchell-Gardner: Melanophycecu 487 though they are more frequently to be found on the pedicels up to fifteen cells long:. At times sessile forms are to be found on the erect filaments about two-thirds of the way up to the apex, but more gen- erally they are terminal on small lateral ramuli on the ereet filaments arising from the middle to near the apices. Rarely there are small lateral forms on the ramuli. Finally, they may rarely be terminal or intercalary and seriate on the erect filaments although the erect fila- ments are usually piliferous. The terminal forms on pedicels, and particularly those on the main erect filaments, are often quite blunt, even elavate. Others on the prostrate filaments are narrow and sharply attenuated, at times terminated by a sterile pointed filament, in effect intercalary. Many of the gametangia, terminal on the ramuli, are composed of a single series of loculi. Most of the others are two or more seriate. The hair filaments do not seem to be abundant. They arise laterally on the main filaments, usually above the center. The meristem in these hairs is at the base which is surrounded by a sheath similar to that described by Sauvageau (1898, p. 47) for MyrUmema. The ramuli are decidedly curved upwards. The species is similar to Myrionema speciosum from the Faeroes, described by Borgesen (1902, p. 421). Our plant is more slender in all of its parts, has much longer gametaiigia. and possesses, in addition to the sessile, secund gametangia figured by Borgesen, occasional intercalary ones and many terminal ones on the numerous clustered ramuli. These ramuli are in turn occasionally branched, usually arise in groups, sometimes whorled, two or three arising from the same cell, in other instances they are secund. The species often grows in association with other members of the Myrionemataceae and the Ectocarpaceae. As a rule, the main erect filaments do not enter into the formation of gametangia, but give rise either to sessile gametangia or to the ramuli. No zoospores have been observed in the so-called zoosporangia. We suspect that these may represent pathological conditions. 18. Compsonema immixtum S. and G. Thallus inconspicuous, the basal filaments creeping among the gametangia of the host ; erect filaments very numerous, all bearing gametangia ; hairs and zoosporangia unknown ; gametangia narrowly ellipsoidal, 24-30/t long. Growing on Colpomerria sinuosa f. deformans. Isla Partida, Gulf of California. 4SS University of California Publications in Botany L VoL - s Setchell and Gardner, Mar. Alg. Gulf, of Calif., 1924, p. 724, pi. 17, fig. 49. This organism is on the border between Myrionema and Comp- sonema as we interpret these two genera (Setchell and Gardner, Phyc. Cont. II-VT, 1922). Under the present condition of the host it is impossible to learn much of the basal portion. It appears that the basal layer was largely developed before the gametangia of the host started to develop. Later the latter pushed out among the basal fila- ments and leaving them behind developed along with the gametangia of the epiphyte. The gametangia project beyond the general surface of the host nearly their entire length. On account of the pluriseriate nature of the gametangia, we have placed it in the genus Compsonema rather than in Myrionema. 8. Hecatonema Sauv. Fronds starting from a single cell, soon developing a series of closely crowded filaments radiating in all directions, forming a com- pact disk ; the branching of radiating filaments often subterminal ; the cells of the radiating filaments wholly or in part dividing once hori- zontally, thus creating a distromatic layer, the lower cells of which may or may not produce short, penetrating, subulate rhizoids; the upper layer of cells producing erect assimilating filaments, wholly or in part producing gametangia, and hyaline hairs; gametangia pluriseriate. Sauvageau, Sur quelq. Myrionem., 1897, p. 248 (p. 88, Repr.). The type species of the genus is H. maculans, collected at Cher- bourg, France, growing on Rhodymenia palmata Grev. The specific name was taken from Phycocelis maculans Collins (Bull. Tor. Bot. Club., vol. 23. 1896, p. 459), the plants of which were distributed in Collins, Holden and Setchell's Phycotheca Boreali-Americana (Exsicc), no. 274. Key to the Species 1. All of the erect filaments transformed into gametangia ...1. H. Lawsonii (p. 489) 1. Part of the erect filaments transformed into gametangia 2 2. Erect sterile filaments clavate 3. H. clavatum (p. 491) 2. Erect sterile filaments cylindrical 2. H. variabile (p. 490) 1925] Setchell-Gardner: Melanophyceae 489 1. Hecatonema Lawsonii S. and G. Plate 53, figs. 5-7 Fronds forming microscopic cushions, 250-500/* diam., irregular in outline; prostrate portion composed of very much contorted, pro- fusely branched filaments with subterminal forking; erect filaments forming a. compact, palisade-like stratum interspersed with numerous hair filaments, 0.75-1.5 mm. long; cells of creeping filaments dividing radially as well as tangentially, forming a pseudoparenchymatous tissue in the center of the thallus, irregular in shape, 4-5/*, diam. ; cells of the hair filaments quadrate at the base and surrounded by a sheath, up to 25 times as long as the diameter above ; zoosporangia unknown ; gametangia cylindrical to slightly fusiform, 25-30/* long, 5-6.5/x broad ; many loculi formed by longitudinal and by oblique walls. Growing on the sporophylls of Nereocystis Luetkeana. Uyak Bay, Alaska. Setchell and Gardner, Phyc. Cont. IV. 1922, p. 379, pi. 40, figs. 5-7. Myrionema vvlgare Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 924 (not of Thuret). Myrionema strangulans Setchell and Gardner. Alg. N.W. Amer., 1903, p. 249 (not of Greville). This plant superficially resembles very closely Myrionema foecun- dum f. simplirissimum 8. and G. Microscopic examination, however, reveals several important differences. The creeping filaments are unique. The branches are very numerous, come off at wide angles and seem, for the most part at least, to be subterminal, or if the terminal cell splits, one of the dichotomy very frequently fails to develop till much later. Many of the cells in the center, and even toward the periphery, divide radially and perpendicular to the surface of the host, forming a pseudoparenchymatous layer. Thus the species, strictly speaking, cannot be said to be distromatic, since the distro- matie condition of a frond is really brought about by radial divisions of the cells of the filaments, but parallel to the surface of the host. The character of the gametangia and of the creeping filaments will not permit of its being placed with Myrionema vulgare Thuret as further described and figured by Sauvageau (1897, p. 186, et seq.). On account of the radial divisions of the creeping filaments and the prevailing biseriate condition of most of the gametangia, we have placed this species with the genus Hecatonema. 490 University of California Publications in Botany [Vol. 8 2. Hecatonema variabile S. and G. Fronds forming circular cushions, 4-7 mm. diam.; prostrate por- tion composed of regularly radiating filaments with subterminal branching, attached by numerous short rhizoids, 1-3 cells long ; erect filaments not densely crowded, not arising successively from the center toward the periphery, unbranched, 400-500** long; hairs unknown; cells of creeping filaments irregular in form, 7-9/* diam.. 1-3 times as long a.s at the margin; cells of erect filaments 4.5-5/* diam. at the base and 2-5 times as long, 8-10/* diam. and 1-3 times as long toward the distal end; zoosporangia (?) broadly clavate, sessile or on short pedicels on the prostrate filaments, rarely lateral or terminal on the erect filaments, 50-65/* long, 20-24/* broad; gametangia cylindrical, tapering slightly at the apices, sessile or on short pedicels on the prostrate filaments or rarely terminal on the erect filaments, 70-120/* long, 9-12/* broad ; loculi 1-2-seriate. Growing on the pneumatocysts of Nereocystis Luetkeana. Carmel Bay, Monterey County, California. Setchell and Gardner, Phyc. Cont. IV, 1922, p. 377. pi. 41. figs. 1-12. From the standpoint of the structure of the gametangia. the type of this species of Hecatonema could scarcely be separated from a Myrionema such as M. phyllophUum S. and G. They, in part, possess uniseriate loculi, but have many perpendicular and slanting walls, making a partial biseriate condition. The basal layer, particularly in the center of the thallus, is distinctly distromatic, developing numerous pegdike rhizoids, which serve to anchor it more firmly to the host, and. for the most part at least, the branching of the filaments is subterminal. whereas that of a true Myrionema is terminal and brought about by the splitting of the apical cell. The size of the plants as a whole varies considerably. The material of the type varies in this respect from 4 mm. to 7 mm. in diameter. In some collections of specimens which we have included under the species, the specimens are as small as 2 mm. in diameter, in others they are up to 10 mm. in diameter. The length of the erect filaments varies from 150/*, to 500/*. Mature gametangia vary from 55/* to 120/* long. The relative number of gametangia terminal on the erect filaments, as well as their size and complexity, is exceedingly variable in different collections. Two or three cells only may meta- morphose in some filaments, in others eight or ten. Occasionally small I92.i| Set eh i ll-Giwdm r: Melanophyceae 491 lateral outgrowths from the metamorphosed cells may occur. The number of erect filaments in proportion to the number of gametangia is a character which seems to be very unstable. In some collections they exceed the number of gametangia almost two to one, in others they do not occur more than in proportion of one to five. They are very irregular in origin along the radiating creeping filaments. Some near the margin of the frond may be nearing maturity before others near the center of the frond have started. 3. Hecatonema clavatum S. and G. Plate 53, figs. 1-4 Fronds forming circular cushions 2-3 mm. diam., with relatively wide margin free from erect filaments; prostrate portion attached by numerous, short, subulate rhizoids, and composed of regularly radiating, closely crowded filaments; erect filaments mostly clavate, unbranched, 190-210//. long; cells of creeping filaments 5-6/x diam., 1.5-2 times as long; cells of erect filaments cylindrical below and 5. 5-6.5//. diam., 2-2.5 times as long, slightly constricted at the dissepi- ments above and 8-11//. diam., 1.5 times as long; zoosporangia (?) sessile or subsessile on creeping filaments, clavate, 60-80/a long, 10-1 4/i broad ; gametangia cylindrical to slightly fusiform, sessile or on short pedicels on the creeping filaments, 55-85//. (up to 111//.) long, 8-12//. broad ; locnli mostly 2-seriate. Growing on the pneumatocysts of Nereoeystis Luetkeana. Mouth of Tomales Bay, Marin County, California. Setchell and Gardner, Phyc. Cont. IV, 1922, p. 378, pi. 40, figs. 1-4. The structures designated zoosporangia (?) in the foregoing diag- nosis of this species were found interspersed among the gametangia. No indication of the formation of zoospores has been observed, which leads to the suspicion that they may be abortive gametangia. The terminal cell of the erect filaments is often longer than the cells below it, and in these filaments in which the cells are divided length- wise, producing the clavate condition, the terminal cell degenerates in a similar manner as do the hairs in other species. The longitudinal division of cells in the upper parts of the filaments is undoubtedly connected w T ith the formation of the short, lateral, often secund, game- tangia characteristic of nearly related species in Hecatonema and in Compsonema, for in a few instances such gametangia were observed in old specimens in wdiich the gametes had, to a large degree, escaped from the characteristic gametangia. 492 University of California Publications in Botang [Vol.8 9. Microspongium Reinke Fronds forming small, cushion-like, rounded, gelatinous masses of various shapes and sizes, composed of a distromatic basal layer the upper cells of which give rise to erect, compact, cylindrical filaments branched or unbranched, among which are intermixed a few hyaline hairs; reproduction by zoospores formed in single (or in seriate?) zoosporangia and by gametes formed in plurilocular gametangia. Reinke, Die braunen Algen der Kieler Bucht, 1888, pp. 16 and 20, Atlas, 1889, p. 11, pis. 7, 8, Algenfl. westl. Ostsee, 1889a, p. 46. Hapalospongidion Saunders, New and little known brown Alg., 1899, p. 37. The type species is Microspongium geiatinosum Reinke. In our judgment there are no fundamental differences between Microspong- ium Reinke (1888, p. 20) and Hapalospongidion Saunders (1899, p. 37), consequently we are adopting the earlier generic name. Both genera form diminutive gelatinous cushions and have a distromatic basal disk which gives rise to erect ' ' hairs ' ' and filaments bearing chro- matophores. Unilocular zoosporangia and plurilocular gametangia are present in both, although there are specific differences. In the first place, the erect filaments in Hapalospongidion geiatinosum are simple. In the second place, the suspected zoosporangia are seriate and inter- calary, probably always so since we doubt Saunder's statement of their being terminal (in complete filaments). In the third place, the game- tangia are pluriseriate and always intercalary, although there may be, at times, only one sterile, enlarged cell with dense contents at the very apex of the gametangial filament. The arrangement and char- acter of both zoosporangia (if normal?) and of the gametangia resemble those of Pylaiella, but we find similar structures occasional among the Myrionemataceae, but generally on plants showing the terminal types of gametangia. It is for these reasons that it seems best to merge Hapalospongidion in Microspongium. Microspongium Saundersii S. and G. Erect filaments 250-750/* long, embedded in a gelatinous matrix ; cells in the lower part of the filament doliiform, 8-10/x long, 4-5ju, broad, in the center of the filament cylindrical, 24-36/t long, 5-6/x broad, in the distal end, 4-5/a long, 7-10/x broad ; hairs scattered, one or more millimeters long; chromatophores discoid; zoosporangia (?) I 925 ] Setchell-Gardncr : Melanophyeeae 493 terminal and seriate, ellipsoidal; gametangia terminal, with loculi 1-2-seriate, 10-14//. broad. Growing on rocks, forming small, irregular, gelatinous or sponge- like cushions, separate at first, but later many coalescing into a common mass. Pacific Grove, California. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 12. Hapalo- spongidion gelatinosum Saunders, New and little known brown alg., 1899. p. 37, pi. 1; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 534. Microspongium Saundersii differs from M. gelatinosum Reinke in having the erect filaments unbranched, the zoosporangia ( ?) seriate instead of single, the gametangia with loculi 1-2-seriate and inter- calary instead of lateral and uniseriate, and in details of measurements. family 4. RALFSIACEAE kjellman Thallus crustaceous, flat, discoid and rounded or crenate at the margins, or deeply lobed and dissected, consisting of a flattened mono- stromatic, discoid, basal layer with marginal growth, soon becoming few to many layered by horizontal partitions, with hairs, single, or in small groups, in depressions of the upper surface, with or without rhizoids from the lower surface; unilocular zoosporangia, either borne at the base of more or less free multicellular paraphyses, collected into more or less distinct patches or "sori" or terminal and forming naked "sori"; gametangia uniseriate or pluriseriate, terminal or borne on lateral branchlets near the tips of the more or less coalescent vertical rows of the thallus. Kjellman, Handbok Skand. Hafsalg., I, 1890, p. 29 (lim. mut.) ; Stragulariaeeae Stroemfelt, Om Algenveg. vid Islands Kuster, 1886, p. 49. Lithodermataeeae Kjellman, loe. eit., 1890, p. 17. The Ralfsiaceae, in the somewhat extended sense in which we con- sider the application of the name, i.e., as including the Lithoderma- taeeae of Kjellman, is closely related to the Myrionemataceae (where Reinke places Ralfsia and its segregates) and the Elachisteaceae, with- out being conveniently placed under either. The frond seems clearly a further development of that of the simpler ilfT/rwwma-species, retaining the flattened circular habit as a fundamental type. The ••oalescence of the vertical filaments, the greater or less separation of fertile areas from sterile, the position of the sporangia and gametangia as to whether terminal or lateral, find close resemblances among the Myrionemataceae or may readily be considered as progressive develop- 494 University of California Publications in Botany [ VoL - 8 ments from them. It does not seem desirable to separate the Litho- derma series of genera and species from the Ralfsia series since both may be considered as derived from the Myrionemataceae, the sporangia being transformed vertical filaments in the Ralfsia series and only the transformed terminal cells in the Lithoderma series, while the game- tangia may be either terminal or on short lateral branches in the Lithoderma series, but terminal in the Ralfsia series, so far as known. Key to the Genera 1 . Zoosporangia included, borne at the bases of the pa raphyses 2 1. Zoosporangia superficial, terminal on the erect filaments.. 12. Lithoderma (p. 500) 2. Thallus entire, or with rounded simple lobes 10. Ralfsia (p. 494) 2. Thallus deeply and narrowly lobed 11. Hapterophycus (p. 499) 10. Ralfsia Berk. Thallus forming flat expansions of indefinite size and more or less circular outline, at least at first, only slightly and broadly lobed at the margin, never dissected into deep and narrow lobes; either attached firmly to the substratum or later separating largely from it, composed at first of a single basal layer of few to many cells in thickness, but later developing closely packed vertical filaments more or less tightly cohering and covered by a common cuticula, those in fertile areas, "sori," later becoming free and rupturing the cuticula; hairs in minute tufts, scattered, projecting beyond the upper surface of the thallus, generally at least in funnel-shaped depressions; zoosporangia lateral, at the bases of the free filaments, "paraphyses" ; gametangia terminal or subterminal, on the erect filaments or lateral from near their tips. Berkeley, in Engl. Bot., Suppl., vol. 3, 1843, pi. 2866; Stragidariir Stroemfelt, Om Algenveg. vid Islands Kuster, 1886, p. 49. Mesospora Weber van Bosse, Not. sur quelq. gen. nouv., 1910, p. 27. We have taken the genus Ralfsia in the original sense and as includ- ing the genera Stragidaria Stroemfelt and Mesospora Weber van Bosse, relying on the habit chiefly to characterize the genus and assign- ing the variations in thickness, position of the gametangia, and aggre- gation of the zoosporangia to distinguish the species, or at most, the sections of the genus. We have been influenced to this point of view by our experience with the variability in these respects among the species of the Ectocarpaceae and the Myrionemataceae. The type of the genus is Ralfsia fungiformis (Gunn.) S. and G. (Fucus fungi- f or mis Gunn.). 1925 ] Setchell-Gardner: Melanophyceae 495 Most specimens collected show only the unilocular zoosporangia, but although gametangia have been seen in the majority of the species, we have not detected them, as yet, with certainty in any of our west coast specimens. Key to the Species 1. Thallus loosely attached, free at the margin 5. R. fungiformis (p. 499) 1. Thallus firmly attached, not free at the margin 2 2. Thallus up to 2 mm. thick, 12 cm. diam 3 2. Thallus less than 0.5 mm. thick, 5 cm. diam 4 3. Paraphyses up to 130 M long 3. R. verrucosa (p. 497) 3. Paraphyses up to 360 M long 4. R. hesperia (p. 498) 4. Zoosporangia 80-95m long 2. R. californica (p. 497) 4. Zoosporangia 60-70> long 1. R. clavata (p. 495) 1. Ralfsia clavata (Carm.) Crouan Fronds flattened, crustaoeous, circular in outline at first, later becoming more or less lobed, firmly applied to the substratum by the whole lower surface with occasional rhizoids, 2-20 mm. diam., about 150ja thick ; zoosporangia narrowly ovoid to pyriform, distributed over the whole upper surface of the frond except, near the margin, 40-60/x long, 13-18/a broad, borne at the bases of the paraphyses ; gametangia short, terminal on the erect filaments, pluriseriate, paraphyses 80-100/x long. 12-15^ broad, slightly clavate; hairs unknown. Growing on rocks, wood, etc., in the upper sublittoral belt, and in tide pools in the littoral belt. Unalaska Bay, Alaska. Crouan, Alg. mar. du Finistere, Fasc. 1, 1853, no. 56 (excl. specim); Reinke, Atlas Meeresalg., 1889, pi. 6, figs. 14-20, Algen- flora. 1889a, p. 48 ; Kuckuck, Bemerk. zur mar. Algenveg., 1894, p. 244. fig. 14; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 253. Myrio- nema clavatum Harvey, Man. Brit. Alg., 1841, p. 124. Linkia clavata Carmichael, MS., in W. J. Hooker, British Flora, vol. 2, 1833, p. 391. The measurements given above for the zoosporangia are taken from Kuckuck {Joe. cit.) and from De-Toni (1895, p. 313). The material from Unalaska cited above is scarcely mature but seems to come within the range of measurements as given in the diagnosis above. We are unable to harmonize the measurements given by Far low (1881 p. 88) for this species, with the material which we have examined from the Atlantic coasts of America and Europe. His measurements for the zoosporangia are 150-180/* long, and 60-70/x broad. These measure- ments are larger than those given for any other species except for R. hesperia S. and G. 496 University of California Publications in Botany [Vol.8 Kuckuck (Bemerkungen, 1894, p. 244 et seq.) figures and describes "plurilocular sporangia" (gametangia) in R. clavata and in R. Bornetii. These were found on the same individual and often in the same sorus with zoosporangia. The materia/ showing these organs was collected in the winter. The study of the genus made from fresh material and from herbarium material at our disposal, unfortunately has been made on summer collections only. Study of the winter mate- rial may reveal the presence of this method of reproduction in the species on our coast. Borgesen, however (1902, p. 432), reported hav- ing discovered "plurilocular sporangia" on R. verrucosa collected in May and July. The Linkia clavata of Carmichael was proposed for specimens from Appin on Loch Linnpe in the west of Scotland. Harvey adopted the specific name of Carmichael, but removed the species to the genus Myrionema, without, however, having seen a specimen. The Crouan brothers referred the species to Ralfsia, quoting Harvey and thus establishing the present binomial. The specimen referred by the Crouans to their Ralfsia clavata seems to be R. verrucosa (cf. Reinke, 1889a, p. 48, as well as Crouan, Florule du Finistere, 1867, pi. 26, fig. 168). Farlow (Mar. Alg. N.E., 1881, p. 88) used the correct binomial and states that "American specimens resemble perfectly the No. 56 of Crouan 's Algues marines du Finistere," but, as noted above, he gives dimensions for the sporangia which are twice as long and several times broader than those which we have assumed to belong to the type of the species. No. 418 of the Phycotheca Boreali-Americana, collected at Spectacle Island, Maine, seems perfectly in agreement with European material from Reinbold and Kylin and the mature sporangia in all average about 60/x high and 18/* in diameter. Our Alaskan specimens agree with these in every detail. We have seen no New England specimens with sporangia more nearly approximating the dimensions given by Farlow than the European and Alaskan specimens, although an occasional sporangium may be as long as 90/x, but such is a very rare occurrence. 1925 J Setehell-Gardner: Melanophyceae 497 2. Ralfsia californica S. and G. Plate 36, fig. 22 Thallus flattened, crustaceous, circular to irregular in outline, adhering firmly without rhizoids to the substratum by the whole lower surface, 280-350/* thick, 2-4 cm. diam., dark brown in color; basal parenchymatous layer composed of 8-12 layers of cells in vertical rows ; zoosporangia numerous, clavate, borne laterally at the base of the paraphyses, distributed uniformly over the central and larger part of the frond, 80-95/* (up to 140/*) long, 16-22/* (up to 32/*) thick; gametangia unknown; paraphyses cylindrical to slightly clavate, 180-220/* long, 9-11/* broad at the outer end ; hairs unknown. Usually found growing on smooth, flattened pebbles and on rock ledges in the upper sublittoral belt, and in tide pools in the littoral belt. Central California. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 2. Ralfsia californica is very similar and is probably closely related to R. clavata (Carm.) Crouan and has passed under that name locally for some time. Careful comparison, however, with material of that species from the Atlantic coasts of America and Europe shows several differences in detail, sufficient to entitle it to specific rank. The speci- mens are, in general, much larger in diameter and are thicker. The paraphyses are generally longer and narrower. The zoosporangia are longer and relatively narrow. The cells of the basal portion are slightly smaller. 3. Ralfsia verrucosa (Aresch.) J. Ag. Frond coriaceous, adhering firmly throughout to the substratum, juvenile state circular in outline, becoming more or less distorted, especially on coming in contact with each other and with other crus- taceous algae, 3-12 cm. diam., 1-2 mm. thick, with concentric zones and radiating lines; color olive brown; erect vegetative filaments curved towards the zone of active growth; paraphyses up to 130/* high, composed of 6-9 cells, the lower 2-3 times as long as broad; zoosporangia ovoid to pyriform, 65-80/* long, 15-25/* broad; game- tangia short, terminal; both reproductive organs on the same indi- vidual, disposed in sori. 4i)S University of California Publications in Botany [Vol.8 Growing on rocks, shells, etc., in the middle and lower littoral belts. Reported from various localities from the Bay of TJnalaska, Alaska, to Monterey Bay, California. J. Agardh, Sp. Alg., vol. 1, 1848, p. 62 ; Reinke, Atlas Meeresalg., 1889, pis. 5, 6, Algenflora, 1889«, p. 48; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 253; Collins, Holden and Setchell, Phyc. Bor.- Amer. (Exsicc), no. LV (?). Cruoria verrucosa Areschoug, in Linnaea, 1843, p. 264, pi. 9, figs. 5, 6. The measurements given for this species by different authors are quite variable. The dimensions of the zoosporangia given in our diagnosis above are taken from Kuckuck (1894, p. 244) and our speci- mens agree ; these were taken as representing the average typical size. Farlow (1881, p. 87) gives the measurements of the zoosporangia as 38/* long and 19/i wide. De-Toni (1895, p. 312) gives 65-80-100,,. long, and 15-30/1 wide. The material quoted from Unalaska is all sterile. Careful comparison with mature material from the eastern and European coasts may show that our California specimens are of a distinct species. The material distributed from southern California under B. ver- rucosa in Collins, Holden and Setchell, Phycotheca Boreali-Americana, no. LV, is very young and the rows of cells are mostly perpendicular to the substratum instead of being curved as is the case with the typical B. verrucosa. It seems quite possible that this plant belongs to some other species, or possibly to another genus. 4. Ralfsia hesperia S. and G. Thallus flattened, coriaceous, circular in outline, adhering firmly to the substratum by the whole lower surface, at times developing rhizoids, 3-4 cm. diam., 1-1.5 mm. thick, with concentric zonation ; color light yellowish brown, dark on drying; paraphyses 290-360/* long, very slightly clavate, 9-12 cells long, 8-H/i broad at the apices, densely crowded with chromatophores above, but few below ; cells of paraphyses 7-9 times as long as broad below, 2-3 times above ; sori extensive, often confluent ; zoosporangia broadly clavate, 120-140/1 (up to 180/x) long, 28-34/1 broad; gametangia unknown. Growing on rocks in the upper littoral belt. Central California. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 2. Balfsia hesperia is very closely related to B. verrucosa in morpho- logical characters. The plants are, as a rule, smaller in diameter and 192 °] Setchell-Gardner: Melanophyceae 499 larger in thickness, are more spongy, lighter colored and have very decidedly larger zoosporangia and longer paraphyses than that species. The exact southern limits of either of these species has not been determined. 5. Ralfsia fungiformis (Gunn.) S. and G. Thallus licheniform, coriaceous, dark brown in color, black on drying, loosely attached in the center by numerous multicellular rhizoids, free around the margin, with free overlapping lobes seemingly from the top of the thallus as well as the margin, circular to irregular in outline, with concentric zones and radiating striae, 2-6 cm. diam., 250-300/* thick; cells forming rows curving upward and downward from a central layer in the lobes ; zoosporangia, gametaiigia and hairs unknown. Growing on rocks in the lower littoral belt. Port Clarence to Sitka, Alaska. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 11. Fueus fungi- formis Gunnerus, Fl. Xorv., 1772, p. 107; (Eder in Flora Dan., 1770, pi. 420. Ralfsia deusta J. Agardh. Sp. Alg., 1848, p. 63 ; Setchell and Gardner. Alg. N.W. Amer., 1903, p. 253 ; Saunders, Alg. Harriman Exp., 1901, p. 421. Zonaria deusta Agardh, Syn. Alg., 1817, p. 40. Padina deusta Postels and Ruprecht, Illus. Alg., 1840, p. 20. The method of reproduction in this species has, apparently, never been observed. It is a plant of the colder waters and has not been collected and studied as much as the other species of the genus. All of the material collected on our coast has been taken in midsummer. In habit and structure, our Alaskan specimens agree with those from northern New England (Phyc. Bor.-Amer., no. 419 and Phyc. Univ., no. 418). 11. Hapterophycus S. and G. Thallus decumbent, dorsiventral, deeply separated into linear lobes radiating from the center, fleshy, short, several times dichotomously branched, concave on the ventral side, convex above, with crenate margins, rounded sinuses and wide, blunt ends; growth apical; central part of the frond consisting of large horizontal hyaline, more or less cylindrical cells, from which rows of hyaline cells curve upward and downward terminating in the small surface cells which are filled 500 University of California Publications in Botany [Vol. 8 with chromatophores above, hyaline below ; cryptostomata with fas- cicles of included hairs scattered promiscuously over the upper sur- face; reproduction by zoosporangia, arising from the same basal cell as the multicellular paraphyses; gametangia unknown. Setchell and Gardner, in Setchell, Alg. Novae I, 1912, p. 233, pi. 25, figs. 1-4. Hapterophycus canaliculars S. and G. Plate 41, fig. 54, and plate 77 Fronds up to 5 cm. wide ; color yellowish brown ; sporangia 65-75/* long, 18-25/a wide ; paraphyses cylindrical, composed of 4-7 cells, 6-7 fi wide, 12-16/a long. Growing on rocks in small tide pools, or on the horizontal surface of flat rocks subjected to wave action, in the upper littoral belt, or even above high-tide limit. San Pedro, California. Setchell and Gardner, in Setchell, toe. cit. Hapterophycus, up to the present, is a monotypic genus, canali- cidatus being the only known species. It has been found only in the vicinity of San Pedro, California, but doubtless may extend much farther south. In its histological structure, it resembles Balfsia deusta, especially in the free lobes of that species. Since the publication of the original description, it has been discovered that on the upper surface there are numerous, small, inconspicuous cryptostomata containing fascicles of hairs which do not extrude beyond the surface of the frond. 12. Lithoderma Aresch. Thallus consisting of a thin parenchymatous disk, circular or irre- gular in outline, of indefinite expansion, attached to the substratum by the entire under surface without rhizoids ; hairs absent ; growth in diameter marginal, and in thickness by horizontal cell division, the cells arranged more or less regularly in vertical rows; reproduction by zoosporangia and by gametangia borne in sori on different individuals, the zoosporangia on the outer ends of the vertical rows of cells, the gametangia either terminal on the vertical rows of cells or lateral on special filaments. Areschoug, Observ. Phyc, part 3, 1875, p. 22. The type selected by Areschoug to represent his newly proposed genus is Lithoderma fatiscens, which is the Ralfsia extensa of Crouan 1925] Setchell-danhu r: M< la noph yceae 501 (Flor. du. Fin., p. 166). It is credited only to the extreme northern part oi' our region. All of the specimens previously reported within our territory are sterile and there is the greatest doubt as to their identity, both generic and specific. Two species have been assigned to our coast, but it is doubtful whether they are correctly referred. We can only list these two species with description and comment. Key to the Species Thallus nearly circular in outline, up to 3 dm. diam 1. L. fatiscens (p. 501) Thallus irregularly elongated or almost linear in outline 2. L. lignicola (p. 501) 1. Lithoderma fatiscens Aresch. Thallus broadly expanded, nearly circular in outline, up to 3 dm. diam., scarcely 1 mm. thick, smooth, shining when wet; the basal layers of cells minute, and somewhat polyhedral, vertical filaments 12-18/x diam., 8-12 cells long; gametangia small, cylindrical or oval, formed by transformation of terminal cells or borne on short, almost colorless branches from the terminal cells; zoosporangia terminal, sessile on the vertical rows of cells, ellipsoid or ovoid. Growing on stones in the lower littoral and upper sublittoral belts. Port Clarence, Alaska, and Iliuliuk, Unalaska. Areschoug, Observ. Phyc. part 3, 1875, p. 22 ; Kjellman, Om Bering. Alg., 1889, p. 49; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 253. The plants of Kjellman from Port Clarence are sterile and those of Setchell and Lawson from Iliuliuk are not to be found for reexamination. 2. Lithoderma lignicola Kjellm. Thallus forming thin, uneven, elongated, almost linear, somewhat confluent crusts, 250/i or more thick ; the basal layer composed of united, radiating branched rows of thick-walled cells, nearly parallel, not flabellate ; the basal layer giving rise to vertical rows of cells, 20 or more, branching occasionally, loosely united above; cells 10-15/a long, 8-10/* diam. ; character of reproductive organs unknown. Growing on piles in the littoral belt. St. Michael, Alaska?. Kjellman, Alg., Arc. Sea, 1883, p. 256, pi. 26, figs. 8-11 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 253. This species was discovered by Kjellman "in the Norwegian Polar Sea." Specimens were taken at Talvik in the middle of September, 502 University of California Publications in Botany [Vol. 8 as described by Kjellman (loc. <■!!.). The specimens were all sterile. We have seen no data as to the reproductive characters of the species since. A single collection of sterile material taken by one of us (Setchell) at St. Michael, Alaska, in July, 1899, was referred by us (loc. ril.) to tli is species. A reexamination of the material and com- parison with material of the type collection, demonstrate that the, Alaskan plants are different, but without assisting in definitely placing them. family 5. ELACHISTEACEAE kjellman Thallus pulvinate or penicillate, more or less lubricous, composed of loosely to compactly intertwined, more or less monosiphonous fila- ments at times agglutinating or anastomosing and forming a pseudo- parenchymatous base, often penetrating the host below and always giving rise above to numerous free assimilating filaments, either zoosporangia or gametangia or both and at times colorless hairs ; game- tangia uniseriate or interruptedly pauciseriate, occasionally lateral, single or grouped; "paraphyses" present or absent. Kjellman, Handbok Skand, Hafsalgfl., I, 1890, p. 41 ; Folsie, Contr. knowl. mar. alg. Norway, I, 1890, p. 92. Elachisteae Reinke, Algenfl. westl. Ostsee 1889a, p. 49. The family of the Elachisteaceae, as separated and indicated by recent writers, includes a number of pulvinate species, all small, but some even minute, largely epiphytes, or possibly growing at times on rocks. They are to be distinguished by their basal portions. In Elachistea, the basal portion is solid due to the agglomeration, or anastomosing of the filaments in this portion of the thallus. In Halothrix, Giraudia, Symphoricoccus, Leptonema, etc., the basal por- tions of the filaments are not united to any considerable extent or at all, but are intertwined and woven together to form a spongy mass. We are inclined to refer to this family C ' ylindrocarpus microscopicus Crouan (cf. Kuckuck, 1899, pp. 49-55, figs. 1-5) as being a link between the Ectocarpaceae and the Elachisteaceae but, because of its pulvinate habit and densely interwoven filaments, approaching the Elachisteaceae more nearly than typical Ectocarpaceae. Several species of Ectocarpus and especially E. hemispheric!** Saund. seem to foreshadow the Elachisteaceae, but are not sufficiently compacted or agglutinated at the base to properly be included in this family. The method of origin of the tufts of the members of this family does not seem to have been traced, but prostrate filaments, either free or 1925 1 Setchell-Gardner: Mekmophyceae 503 more or less united into a disk or irregular membrane probably pre- cede the development of the compacted or agglutinated erecl filaments, thus assisting in linking the Elachisteaceae with the Ectocarpaceae and the more pulvinate Myrionemataceae. Key to the Genera Basal cells of the thallus more or less firmly agglutinated; paraphyses present 13. Elachistea (p. 503) Basal cells of the thallus loosely interwoven; paraphyses wanting 14. Gonodia (p. 505) 13. Elachistea Duby Frond tufted or expanded, composed of a colorless basal pseudo- parenchymatous mass of cells, which may or may not penetrate the host, and arising from these, closely compacted ''paraphyses," long free assimilating filaments and numerous long hairs, present only in certain species; reproduction by zoosporangia and by gametangia, both borne among the "paraphyses." Duby, Bot. Gall., 1830, p. 972. We have adopted the original spelling of Duby who founded the genus on the Conferva scutulata Engl. Bot., pi. 2311. The type species of the genus, E. scutulata, was collected in 1811 by Mr. Borrer at Brighthelmston on the coast of Sussex, England. There seem to be two species within our territory, both occurring to the north of the California coast. Opospermum of Rafmesque (Precis decouv. et trav. somiol., 1814) seems, from the description, to be unidentifiable, at least by no means certainly to apply to any member of Elachistea. Key to the Species 1. Paraphyses decidedly curved, strongly moniliform, with upper cells as long or longer than broad 1. E. fucicola (p. 503) 1. Paraphyses straight or slightly curved, barely constricted at the dissepiments, with upper cells shorter than broad 2. E. lubrica (p. 504) 1. Elachistea fucicola (Velley) Aresch. Plate 38, figs. 33-35 Thallus fasciculate, 4-7 mm. high, the basal portion composed of a small, usually hemispherical, dense, lubricous mass of closely inter- twined branched filaments, giving rise to numerous free paraphyses, sporangia and assimilating filaments; paraphyses clavate, uncinate, moniliform, 170-200^ long, 18-24/a diam. above, 1-W/x diam. below; 504 University of California Publications in Botany [Vol. s assimilating filaments briefly tapering near the base, and long and gradual tapering above, variable in size, 20-70/*, mostly 40-50/*, diam. not ending in a hair; cells ().f>-2 times as long, cell wall thick; chro- matophores in actively growing cells a network extending throughout the cell, later becoming granular; zoosporangia narrowly cuneate, rounded above. 90-110/* long, 22-26/* diam., 200 or more spores escap- ing at the apex in a utricle; gametangia unknown. Growing on different species of Fucus in the lower littoral belt. Sitka. Alaska, to central Oregon (Coos Bay). Areschoug, w? Linnaea, vol. 16, 1842, p. 235, pi. 8, figs. 6, 7. Conferva fucicola Yelley, Mar. PI, 1795, no. 1. This species was first reported as being in our region by Ruprecht in 1851 (p. 389) from Sitka. No subsequent report of its having been collected had come to us until one of us (Gardner) collected it in 1!)17 at the same station. There seems to be some question as to the exact limits of Elachistea fucicola, but it has appeared most reasonable to us to adopt, as typical, the plant figured by Kuetzing in the Tabulae Phycologicae (vol. 7, pi. '.)y)). We find plants within our territory corresponding in general to Kuetzing 's figure and epiphytic on species of Fucus. The free filaments are long attenuate at both the tip and the base. They vary decidedly in diameter and in length and breadth of cells even in the same individual, and the paraphyses are decidedly curved and de- cidedly moniliform, with the uppermost cells usually as long as, or longer than, broad. Our plants agree fairly well, also, with no. 417 of the Phykotheka Universalis collected at Warnemunde on the north coast of Germany by Heiden. Our figure (plate 38, figs. 33-35) gives the essential characteristics of structure. Elachistea fucicola is usually, at least, more olivaceous than E. luorica and all parts of the plants are firmer and collapse less on drying than do those of E. lubrica, Thus far we have detected E. fucicola only on species of Fucus. 2. Elachistea lubrica Rupr. Thallus usually forming distinctly separate tufts, 3 mm.— 2 cm. (9-11 mm. Rupr.) high, lubricous; mass of paraphyses relatively small and loosely intertwined ; hairs absent ; erect free filaments abruptly attenuated at the base, and gradually attenuated above the center, 18-22/* diam. at the base, 38-48/* in the widest part, tapering to 9/i or less at the apex ; cells of erect filaments cylindrical to slightly l925 l Setch&ll-Gardn&r: Mekmophyceae . 505 doliiform, 2-3 times as long as the diameter at the base, 0.3-0.8 times as long- a short distance above the base. 1-1.5 times as long in 1 1n- widest part, up to 6 times as long at the apex ; chromatophore a fine nodular network in the young cells, separating more or less into spheroidal masses in the older cells; zoosporangia subovate to broadly clavate. 75-100/* (up to 135/*) long, 30-40/* broad at the outer end; gametangia unknown; paraphyses clavate, about 225/* long, slightly curved at times but not sensibly constricted at the dissepiments. Growing on other species of algae, especially those of Rfvodymenia. From Prince William Sound to Wrangell, Alaska. Ruprecht, Tange Ochots. Meer., 1851, p. 196 (388 Orig.) ; Setchell and Gardner. Alg. N.W. Amer, 1903, p. 248; Collins, Holden and Setchell. Phyc. Bor.-Amer. (Exsicc), no. 828. The type of Elachistea lubrica Rupr. came from the Ochotsk Sea, but Ruprecht states that specimens from Russian Lapland are in close agreement. The most typical specimens seem to grow upon species of Halosaccion and Rhodymenia. Specimens from Yakutat Bay collected by Saunders (Phyc. Bor.-Amer., no. 828) and by Albin Johnson as well as at Orca by Setchell, seem to agree well with speci- mens distributed from Spitsbergen by Kjellman and from Mehavn in Finmarken (Phyk. Univ., no. 366) collected by Foslie. The free fila- ments taper both above and below as in E. fucicola, but the paraphyses are less curved, only slightly, if at all, constricted at the dissepiments, and the upper cells are inclined to be broader than high. 14. Gonodia Xieuwland Fronds minute, tufted, composed of a mass of colorless, densely intertwined filaments, mostly penetrating the host, giving rise to a dense mass of erect assimilating filaments and "hairs," at least at times; ' paraphyses " wanting; reproduction by zoosporangia and gametangia. Xieuwland, Critical Notes IX, 1917, p. 30 ; Setchell and Gardner, .Alar. Alg. Gulf Calif., 1924, p. 722 ; Myriactis Kuetzing, Phyc. gen.. 1843, p. 330. The distinctions between Gonodia and Elachistea are not always readily discernible in all of the species. In the younger stages of some of the species of Elachistea, and at times in the older stages, the basal portion, especially if it penetrates the host, is of about the same degree of compactness as it is in the mature stages of the species of Gonodia. ;,0(i University of California Publications in Botany [Vol. 8 Typically, however, in the mature stages of the two genera, Elaehistea has a firm, closely intertwined, agglutinated mass of filaments for the basal port ion, and Gonodia remains free and loosely intertwined. We are using this character as a basis of separating the two genera. Key to the Species 1. Basal cells of the erect free filaments symmetrical 1. G. Johnstonii (p. 506) 1. Basal cells of the erect free filaments asymmetrical... 2. G. Marchantae (p. 506) 1. Gonodia Johnstonii S. and G. Plants forming dense minute tufts in the conceptacles and on other parts of the host, 160-200/* high, attached by rather deeply penetrat- ing, sparsely branched, rhizoidal filaments; erect fronds forked at the surface of the host, vegetative filaments unbranched above, decidedly clavate and blunt, at times tapering upwards above the center ; cells in widest part 10-14/* diam., 1-1.5 times as long as broad, slightly constricted at the dissepiments; zoosporangia broadly clavate, 65-75/* long, 18-22/1 broad ; gametangia cylindrical, 60-75/* long, 6-9/* broad, densely fasciculate ; both sets of reproductive organs borne on the same plant at the surface of the host. Growing on Sargasswm msulare. San Marcos Island, Gulf of California. Setchell and Gardner, Mar. Alg. Gulf Calif., 1924, p. 722, pi. 17, figs. 46, 47. *to K 2. Gonodia Marchantae S. and G. Fronds forming dense tufts with a dense mass of colorless filaments at the base penetrating the host, the free portion about 200/* long ; fila- ments unbranched above the host, the lower portion composed of 2-3 long, narrow cells, abruptly changing into 2-3 asymmetrical swollen cells, then gradually attenuated upwards to blunt apices; widest cells 18-22/*, the length of the cells in the upper part equaling the breadth ; pseudoparenchymatous cells doliiform to subspherical ; zoosporangia broadly clavate, 55-65/* long, 22-26/* broad; gametangia cylindrical, densely fasciculate 55-65/* long, 6-7/* broad ; both sets of reproductive organs borne on the same plant at the base of the free filaments. Growing on the fronds of Sargassum horridum. La Paz, Lower ( lalifornia. Setchell and Gardner, Mar. Alg. Gulf Calif., 1924, p. 723, pi. 17, fig. 48. I 925 ] Setcheil-Gardner: Melanophycea* 507 G. Marchantae differs from G. Johnstonii in the character of the basal penetrating portion, the former having few narrow filaments and the latter having a dense, copious base. Two or three cells in the lower part of the free filaments are usually asymmetrical in G. Mar- chantae and not in G. Johnstonii. Neither species has hairs nor any indication of having had them, a prominent character as figured by Thuret and Bornet in Etudes Phycologiques, pi. 7, figs. 2-6, for Elachista pnlrinata and mentioned by Yendo as "paraphysibus pau- cioribus" in Myriactis Sa/rgassi (Novae Alg. Japon., 1920, p. 3). G. Marchantae is close to G. pnlrinata in the character of the basal penetrating portion. G. Johnstonii in general resembles G. monili- formis but is much smaller throughout. family 6. LEATHESIACEAE fam. NOV. Thallus thick, carnose, expanded on the substratum and solid or irregularly globular and hollow, arising from a flattened, monostro- matic (?) persistent or evanescent disk; inner cells large, colorless, loosely parenchymatous of di- to trichotomous filaments, outer colored, in anticlinal rows, generally decreasing in size from within outwards, held together, at least loosely, by the surrounding jelly ; zoosporangia and gametangia with loculi uniseriate or nearly so, both immersed among the external anticlinal rows present and borne on the same or similar plants ; epiphytic or saxicolous. Corynophlaeceae Oltmanns, Morpli. und Biol, der Alg., 1922. p. 23. As will be made apparent later, it seems most desirable to restrict the family Chordariaceae to the genus Cliordaria, as typified by C. ftageiliformis, an erect branched plant with distinct elongated axis and lateral branches of limited growth, with subapical terminal meristem, not trichothallic in the same degree as in the iEgiraceae and with only the unilocular zoosporangia thus far known, suggest- ing the strong probability of a microscopic gametophyte. The iEgiraceae are more or less elongated, simple or branched plants, with strongly trichothallic apical meristems, and both zoosporangia and gametangia borne on macroscopic plants. The Corynophlaeaeeae of Oltmanns (1922, p. 23) includes the genera Cylindracarpus, Micro- corijne, Strepsithatia, Conjnophlaea, Myriactis, and, presumably, Leathesia, although Oltmanns might be suspected of merging (toe cit., p. 26) Leathesia (Gray, 1821) in Corynophlaea (Kuetzing, 1843). AVe have referred Myriactis (Gonodia) to the Elaehisteaceae because of its close resemblance to Eiachistea. Cyiindrocarpus micrascopicus 508 University of California Publications in Botany [Vol. 8 Crouar Beems to as to be more properly included in the Ectocarpaceae, as a further developmenl of such forms as Ectocarpus hemisphericus and /•:. chantransioides and on the way toward simpler ^giraceae, such as we conceive may be represented by Microcoryne (cf. Oltmanns, lac. cit, pp. 25, 26). Sin psit Italia is unknown to us, but on account of its gelatinous character may possibly be considered to be a slightly developed, or even degenerate relative of Corynophlaea. Corynophlaea, if it be distinguished from Leathesia, is more generally solid, with longer assimilating filaments which are also less swollen at the tip; Kylin (11)07. p. 83) is of the opinion that the three genera, Micro- coryne, Corynophlaea, and Leathesia are to be retained as representing a series in complexity, leading up to the most highly developed species of Leathesia. As indicated above, Microcoryne, on account of its elon- gated form and greater or less differentiation of layers seems to approach Mgvra and the resemblance in structure to Mgira has been generally noticed and commented upon. Between Leathesia and Cory- nophlaea there seems to be no satisfactory distinction either as between solid and hollow, or between longer and shorter cortical filameiits. We have therefore merged the two genera under Leathesia and designated the family as Leathesiaceae. We can see no resemblances sufficiently close to unite Petrospongium with Cylindrocarpus (considering C. niirroscopicus Crouan as the type). We have therefore kept them distinct for reasons to be given later. The Leathesiaceae are carnose algae of flattened or globular form, solid or hollow, arising from a primitive basal disk, with the inner cells large and colorless, more or less agglutinated, the cortical cells in short filaments with chromatophores, without exserted structures of any kind except colorless hairs, with peripheral growth, and having gametangia and zoosporangia on the same or on similar plants. Key to the Genera 1. Zoosporangia more or less difform, usually attached laterally 15. Petrospongium (p. 508) 2. Zoosporangia pyriform to ellipsoidal, attached at the base.. 16. Leathesia (p. 510) 15. Petrospongium Naeg. Thallus small, thin, flat, usually encrusting rocks, with or without rhizoids, gelatinous, composed of 3 more or less distinct tissues, viz. : (1) a basal layer, monostromatic ( ?), giving off rhizoids, whence arise (2) loosely compacted, di- or trichotomous filaments, decumbent below, curving to erect above, consisting of elongated, cylindrical or swollen. 192.")] SetcheU-Gardner: Melcmophyceae 509 colorless cells with scanty chromatophores, and (:{) a cortical layer of short, branched, closely compacted, erect filaments composed of short, colored cells forming; a more or less smooth or wrinkled surface ; zoosporangia, embedded within the thallus, more or less cylindrical, becoming difform, usually attached laterally ; gametangia unknown. Xaegeli, in Kuetzing, Tab. Phyc, vol. 8, 1858, p. 2, pi. 3, fig. II. Comparison of Petrospongium Berkeleyi (Grew) Naeg. with Cylin~ drocarpus microscopicus Crouan reveals such a difference in habit, structure, etc., that it seems impossible to place the two species in the same genus. We are also of the opinion that they belong in separate families. The figures of Cylmdro carpus microscopicus, especially those of Kuckuck (1899, p. 88, pi. 6, figs. 1-5) show a plant of definite habit, growing in tufts (mucose, fide Crouan) rather tending to elongate although only slightly so. without its individual branched filaments being agglutinated into the definite thallus with inner and outer distinct tissues as in the Leathesiaceae, but united into a spongy (fide Kuckuck) mass by intertwining rhizoidal corticat- ing filaments. We are inclined to place Cylindrocarpus micro- scopicus in the Ectocarpaceae, as an extremely differentiated type of such forms as Eetocarpus liemisphericus and E. chantransioid.es and looking toward the iEgiraeeae and perhaps, also, the Leathe- siaceae, but not properly of them or, to the same degree, differen- tiated into definite inner transparent and outer colored layers. Petrospongium Berkeleyi (Grev.) Naeg. and P. rugosum (Okamura) S. and G., especially the latter, have their tissues differentiated and of agglutinated, level-topped filaments and are, in our opinion, perfectly typical members of the Leathesiaceae. Petrospongium rugosum (Okamura) S. and G. Plate 39, figs. 42, 43 Thallus adhering more or less loosely to the substratum, circular to irregular in outline, flat or thrown up into folds, more or less spongy and lubricous, 2.5-5 cm. (up to 10 cm., cf. Okamura) diam., 1.5-2.5 mm. thick, of a dark, glossy, chestnut brown color ; cells of the cortical layer 8-11//. diam., 1-2 times as long as the diameter ; zoosporangia narrowly ellipsoidal, at times decidedly difform, attached laterally a little below the middle and near the base of the cortical filaments, 75-90/x long, 16-22/x broad. Growing on rocks in the upper littoral belt exposed to surf. Central and Southern California. 510 University of California Publications in Botany L V(JL - 8 Setchell and Gardner, Phyc. Cont, VII, 1924, p. 12. Cylindro- carpus rugosa Okamura, Alg. Japon. Exsicc, no. 88, 1903, Icon, Jap. Alg., vol. 1, 1907, p. 20, pi. 5, figs. 1-6. Petrospongiuni Berkeleyi Collins, Holdcn and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 232, IS!) 6 (not of Naegeli). This enrions, dark colored, rather firm, yet carnose plant is com- mon on rocks exposed to snrf along the central Californian coast and has been found on that of southern California as well. Its peripheral filaments differ in size (about twice the diameter) and in branching from those of the European species and agree closely with those of the Japanese species. 16. Leathesia Gray Fronds irregularly globose, solid in the juvenile stage, soon becom- ing hollow in some species and remaining solid in others, decidedly carnose, interior consisting of radiating, dichotomously or trichotom- ously branched filaments composed of large, irregular, colorless cells toward the base or center as the case may be, merging into smaller cells toward the periphery ; cortex composed of a short series of small, color bearing cells set in a firm mucous matrix; zoosporangia and uniseriate gametangia known, both arising at the base of the cortical filaments; hairs single or in fascicles. Gray, Nat. Ar. Brit. PI., vol. 1, 1821, p. 301. The genus Leathesia, as understood and interpreted by us, should include the Corynophlaea and Corynephora of Kuetzing (1843, p. 331). The chief distinction between Corynophlaea and Leathesia as diag- nosed, seems to be that in Corynophlaea the adult thallus is attached by a single layer of cells from which arise all the other cells, forming a flattened, hemispherical or spherical, solid structure depending upon the nature of the substratum, while in Leathesia the adult thallus is more or less spherical and usually hollow. Structurally and in mode of development there seems to be no fundamental difference. The basal part is practically the same in both genera in the early stages of development (cf. Oltmanns, 1922, p. 26, but cf. also Okamura, 1907, p. 81, pi. 18, figs. 9-14). The cells of the basal portion are more or less modified to serve as attaching cells in both genera, It is difficult to determine in the diminutive species of both genera whether or not all of the specimens are solid or hollow. Kuetzing indeed transferred the type species of his genus Corynophlaea, C. oaltica, to his genus Corynephora in Tabulae Phycologicae (1858, pi. II), which is hollow L925] Setchett-Gardner: Melanophyceae 511 and differs in no fundamental characters from Leathesia, In the larger species of Leathesia the whole under portion sooner or later disintegrates, leaving only the marginal cells for attaching the plant. Until more can be learned of the lifeddstory of and extent of variation in these genera, it seems to us best to combine them and place them all under the oldest name, Leathesia. Key to the Species 1. Fronds diminutive, up to 1.25 mm. diam 1. L. nana (p. 511) 1. Fronds variable in size, up to 8 cm. diam 2 2. Paraphyses terminating in a spherical or subspherical cell 3. L. amplissima (p. 513) 2. Paraphyses clavate, gradually and uniformly enlarging upward 2. L. difformis (p. 511) 1. Leathesia nana S. and G. Plate 43, fig. 67 Thallus solid, spherical, 500-800/x (up to 1.3 mm.) diam.; color dark brown ; cells in the center subspherical, 40-55/* diam. ; paraphyses 30-40/* long, composed of 3-4 cells, terminal cell enlarged, sub- spherical, 10-11/* diam. ; hairs 300-400/* long, 4-5/* diam., scattered, not in fascicles ; gametangia with 4-5/* loculi, uniseriate, 3.5-4.5/* diam. ; zoosporangia unknown. Growing on the leaves of Phyllospadix near the outer ends. Monterey, California. Setchell and Gardner, Phyc. Cont. VII, 1924, p. 3. The host upon which this species of Leathesia was found has a wide distribution along the west coast of the United States and careful observation may show that this little known epiphyte is coexistent with its host. In structure it is not far unlike the Corynephora marina of Kuetzing (1843, p. 331 and 1858, pi. 3, fig. I) but is of smaller dimensions and has much smaller interior cells. 2. Leathesia difformis (L.) Aresch. Plate 40, fig. 52 and plate 43, figs. (55, 66 Fronds gregarious, very variable in size, up to 6 cm. diam., spherical and solid when young, becoming hollow and more or less lobed at maturity, hairs in small scattered groups; cortical filaments clavate. 5-6 cells long, with eidarged, subspherical to spherical terminal cells ; color olive brown ; gametangia cylindrical, 34-45/* long, 4-6/* .">r_' University of California Publications in Botany [Vol. 8 broad, with 6-8 uniseriate loculi; zoosporangia pyriform or ovoid, 35-60ju, long, 16-22/x broad. Growing on other algae and on roeks in the littoral belt. Ranging from Bering Sea, Alaska, to southern California. Areschoug, Phyc. Scand. T, 1846, p. 154 (Repr.) ; Saunders, Alg. Harriman Exp., 1901, p. 423; Setchell and Gardner, Alg. N.W. Amer., 1003, pp. 249, 250; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsice.), no. 829; Tilden, Amer. Alg. (Exsicc.), no. 243 (?). Tre- nt ell a d iff or in is Linnaeus Fl. Suec. (ed. two), 1755, p. 429. Regarding the dimensions of Leathesia d iff or mis as a whole or especially of its parts the literature is practically silent. Areschoug {toe. eit.) states "Thallus nunc semine Cannabidis vix major, nunc 2-5 uncias diametro aequans, " but gives no definite statements regard- ing the size of the various parts. Farlow (1881, p. 82) gives as the diameter of the frond "from half an inch to two inches. " De-Toni (1895, p. 422) states sporangia "35 x 17." It seems to us quite prob- able that this name has been used freely to include a number of species somewhat similar in general appearance, but in reality distinct. We feel that much more work should be done on our Pacific coast material before the limits of the species can definitely be determined. Gardner no. 3462 was collected at Tomales Bav, California, in August and shows mature zoosporangia and gametangia on the same individual. The gametangia have 6-8 loculi. Gardner no. 4501 was collected at Pacific Grove, California, in December, and has only game- tangia. These have 9-13 loculi. It may be that we are dealing with two entities, but for the present we are grouping them under L. dif- formis. Plants from both of these numbers are figured on plate 43. The dimensions given in our diagnosis above are more or less a composite but represent a fair average of the size of the structures found in the plants which we have included in the species. We have examined material of this species in our herbarium from Kylin, col- lected at Strangahnfund and at Varberg, Sweden, which may be considered to be topotypes of L. difformis. The material from both of these localities is bearing fruit and apparently mature. The fol- lowing measurements of their parts will indicate the relation in this respect to our Pacific coast material. Paraphyses, composed of 3-4 cells, 38-44/t long in plants with gametangia, up to 75/*, long in plants with zoosporangia, terminal cells 14-18/a in diameter. Gametangia composed of 4-6 uniseriate loculi, 20-25/* long, 5-6/x in diameter. Zoosporangia are 30-50/* long, 15-22/* in diameter. 1925] Setchcll-Gardner : MeJaiwphijceae 513 3. Leathesia amplissima S. and G. Plate 43, fig. 64 Thallus hollow, subglobose to irregular, attached by a broad area, the central part of which disintegrates on nearing maturity, 3-8 cm. diam. ; color yellowish brown ; interior colorless cells relatively large, compact and angular from compression, becoming radially arranged and smaller, toward the surface ; hairs single, scattered over the whole free surface, 11-13//. diam. ; paraphyses clavate, enlarging gradually from the base to the apex, 55-65/1, long, 4— 7/x diam. at the base, 7-12/t at the apex, 4-5 cells long ; chromatophores densely crowded ; zoo- sporangia ovoid to ellipsoidal, 38-46/x long, 22-28/t broad ; gametangia unknown. Growing on rocks and over other small algae, in tide pools in the lower littoral and upper sublittoral belts. Central California (Pacific Grove and Monterey). Setchell and Gardner, Phyc. Cont., VII, 1924, p. 3. Leathesia amplissima is not readily distinguishable in gross mor- phological characters from L. difformis as we understand that species at present. The histological structure, however, is essentially different as may be seen from our descriptions and figures. family 7. PUNCTARIACEAE kjellm. Fronds from filiform and almost monosiphonous to saccate, ligulate or broadly expanded and membranaceous, attached by a relatively small parenchymatous disk or by rhizoidal filaments, all but the mono- siphonous species differentiated into an inner and an outer tissue, the inner mostly of colorless cells and the outer of color bearing cells; reproduction both asexual, by zoosporangia projecting slightly or not at all beyond the surface, and sexual, by gametangia projecting wholly or in part beyond the surface; both zoosporangia and gametangia transformed single surface cells, either scattered or collected into definite sori ; hairs ( and paraphyses ? ) present in some. Kjellman, in Gleerup, Enum. PL Scand., 1880, p. 9 (fide Foslie, Contr. Knowl. Mar. Alg. Norway, I, 1890, p. 98) ; Oltmanns, Morph. und Biol, der Alg., 1922, p. 49 et seq. Punctarieae Thuret, in Le Jolis, Liste alg. mar. Cherb., 1863, p. 14 (in part) ; Farlow, Mar. Alg. New Eng., 1881, p. 63. As usually limited, the Punctariaeeae include the genera Omphalo- phyllum Rosenv., Phaeosaccwn Farlow, Punctarki Grev., Desmo- 514 University of California Publications in Botany [Vol.8 trichum Kuetz., Pogotrichum Reinke, Litosiphon Harv. and Corycus Kjellm. We know of but one of these in our territory, viz., Punctaria. Structurally the simplest forms of the Punctariaceae, as generally understood, are almost monosiphonous. The more complex forms usually included in the family are either filamentous and solid, saccate or membranaceous. They agree in general in that the zoosporangia and gametangia arise as transformed single surface cells, projecting beyond the surface but slightly, if at all, but if projecting not accom- panied by pluricellular paraphyses. They form fairly definite and circumscribed sori. The Punctariaceae have for their type genus Punctaria Grev. whose type species is P. latifolia Grev. Their relation is closely with Striariaceae, on the one hand, and the Scytosiphonaceae on the other. This relationship is so close that it is largely a matter of convenience to separate them. From the Striariaceae they may be separated by the tendency of their fronds even in the simplest Desmotrichum species, to develop a membrane rather than a cylindrical frond. In case this distinction is made, Pogotrichum and Litosiphon, under such understanding, should be referred rather to the Striariaceae and to the neighborhood of Stictyosiphon. While the species of Desmotrichum and Punctaria are both supposed to have hairs, yet some of our Pacific coast species of Punctaria have no trace of hairs in adult plants. There is an absence of hairs also in the known species of Phaeosaccion and of Omphalophyllum, two genera usually referred to the Punctariaceae. The two genera just mentioned also lack specialized sporangial cells in groups and it may be found desirable to separate them and give them a family designation of their own. Neither of these genera have been detected as yet in our territory, so that we have no opportunities for careful study. Corycus of Kjellman is hollow and, so far as descriptions inform us, devoid of hairs in the adult stages. It is flattened, even if it is hollow, and seems to belong in the Punctariaceae on that account as well as agreeing with both Striariaceae and Punctariaceae in having circumscribed sori which do not project beyond the surface. The Punctariaceae differ from the Scytosiphonaceae in having the sori circumscribed and, commonly, of regular shape. The Scyto- siphonaceae have extensive, indefinite and confluent sori and, thus far, have been detected only with gametangia. 1;| -- 5 ] Setchell-Gardner: Melanophyceae 515 17. Punctaria Grev. Fronds simple or more or less branched, monosiphonous, usually complanate, membranaceous, linear to broadly expanded, attached by a small discoid base, composed of several layers of cuboidal cells similar in size or those of the surface layer much smaller ; reproduc- tion by zoosporangia immersed in the tissue or slightly projecting or bulging beyond the surface, and by gametangia usually partly or wholly immersed in the tissue; both gametangia and zoosporangia representing transformed surface cells, single or in small groups, scattered promiscuously on both sides of the frond ; hairs present in some species; paraphyses wanting. Greville, Alg. Brit., 1830, p. 52 and Syn., p. xlii. Diplostromium Kuetzing, Phyc. Gen., 1843, p. 298. Phycolapathum Kuetzing, Phyc. Gen., 1843, p. 299 (pro parte maxima). Desmotrichum Kuetzing, Phyc. Germ., 1845, p. 244. Homoeostroma J. Agardh, Anal, alg., Cont. Ill, 1896, p. 7. Nematophlaea J. Agardh, loc. cit., 1896, p. 12. Rhadinocladia Schuh, Rhodora, vol. 2, 1900, p. 3, pi. 18. There can be little question as to the idea underlying Greville 's proposal of the genus Punctaria, if we accept his P. latifolia as the type. The specimen examined by Greville in its living state was collected at Sidmouth on the south coast of England by Mrs. Griffiths. It seems sufficiently certain that Wyatt's Algae Damnonienses, no. 9, is practically a topotype, having been collected along the same coast and not far removed from the actual type locality. Wyatt 's specimen answers in every detail to Greville 's description and we base our idea of the genus Punctaria and of the type species, Punctaria latifolia Grev., on it. It is probably the same as the plant figured by Harvey in Phycologia Britannica (plate 8). The Wyatt specimen is a Punctaria also in the sense of J. G. Agardh (1896, p. 4, pi. 1, fig. 1), although a distinct species from any included by him. The Punctaria of Greville, besides including P. latifolia Grev. and P. plantaginea (Roth) Grev., was also extended to include P. tenuissimum (Ag.) Grev. This last species was removed later by Kuetzing to his Diplo- stromium (1843, p. 298) and this genus, in turn, in spite of the refer- ence to it as first or "type" species of Viva pi ant ag mi folia Lyngb., is identical, partly in foundation and partly in content, with Kuetz- ing's later genus Desmotrichum (1845, p. 244). Kuetzing's genus Diplostromium has, then, the characters of Greville 's genus Punctaria. 516 University of California Publications in Botany [Vol. 8 The Punctaria tenuissima Grev. is probably identical with P. undulata J. Ag., now generally referred to Desmotrichum. It is much more difficult to determine the exact status of Punctaria plantaginea (Roth) Grev. The original is the plant figured in Dil- lenius (1741, pi. 9, fig. 4) which, as well as the description, was founded on a plant studied by Micheli. Smith (1810, pi. 2136) again figured it and described it from Micheli 's material and notes which had been communicated to him by Professor 0. Targioni-Tozetti long after the death of Micheli. Whence the original material was derived is uncertain, but it seems likely to have been the Adriatic Sea since Micheli lived at Florence. In such a case, it is probable that it is not the species which Greville referred to his Punctaria plantaginea but nearer to P. latifolia Grev. There seems to be little reason for doubt- ing that the Punctaria plantaginea as Greville understood it is also a Punctaria in the sense of J. G. Agardh (loo. cit.). J. G. Agardh has proposed a segregate from Punctaria, viz., Homoeostroma to receive the plant from Cherbourg, described and figured as Punctaria latifolia by the French phycologists, Thuret and Bornet (1878,p . 13, pi. 5). Whereas both the Punctaria latifolia and the P. plantaginea of Greville have the external layer on each surface of the frond distinctly smaller and more deeply colored than the internal layers, in the P. latifolia of Thuret and Bornet (but not of Greville) the external layers are of the same sized cells as those of the interior and slightly, if at all, more deeply colored. In our experience, however, there are forms difficult of placing in case Punctaria and Homoeostroma are considered separate genera. There remain for consideration the relations of Punctaria and Desmotrichum. Kuetzing (1845, p. 244) founded the latter genus on D. balticum, a slender, yet membranaceous form with hairs borne singly and gametangia conical and practically superficial, at least judging from his figures (Kuetzing, 1843, pi. 20 II, figs. 2, 4, 5-8). Reinke, however, considers Kuetzing 's D. baltimim to be a mono- siphonous species although rarely it may be two to four cells wide, while D. undulatum (J. Ag.) Reinke is the broader species. Aside from the slender character of the frond, the strongly projecting or even superficial gametangia and the hairs borne singly seem to dis- tinguish the genus. We find, however, strongly projecting, even superficial gametangia in broad species with hairs in groups. We also find certain delicate, but broad species with single, promptly deciduous hairs, and we find species in which there is no indication of hairs, 1925 ] Setchell-Gardner: Melanophyceae 517 even traces of deciduous ones. Consequently we do not feel certain of anj- constant difference among' these three genera, viz., Punctaria., Homoeostroma and Desmotrichum. We also do not find any definite distinctions to separate Shuh's Rhadinocladia (1900, pp. Ill, 112) from the simpler Desmotrichums (cf. especially Kylin, 1907, pp. 66-68), nor are we able to distinguish satisfactorily Kuetzing 's Diplo- stromium (1843, p. 298) while J. G. Agardh's Nematophloea (1896, p. 12), if Kuetzing's figure (1856, pi. 45, I) is to be considered typical, seems to have nothing to distinguish it from Kuetzing 's Phycolapathum (excluding synonymy of type species, pro parte) and Punctaria. Phycolapathum Kuetzing (1843, p. 299, pi. 24, II) is founded on Greville's Punctaria latifolia, although Laminaria debilis Ag. is quoted, possibly erroneously, as a synonym. The plate of Kuetzing represents a Punctaria. We have, therefore adopted the name Punc- taria for all of our species, but we have seen none of the slender Desmotrichum type on our coast, Punctaria has received the sanction of the International Botanical Congress (Rules, 1912, p. 76) for the generic name. Key to the Species 1. Fronds with numerous marginal crenulate lobes 2 1. Fronds with entire, smooth or undulate or coarsely ruffled margins 3 2 Fronds large, up to 50 cm. long, chartaceous, composed of 6-7 layers of cells 6. P. chartacea (p. 521) 2. Fronds small, 10 cm. long, composed of 4 layers of cells. .4. P. lobata (p. 519) 3. Fronds broadly expanded, up to 30 cm. broad, very fragile. .7. P. expansa (p. 521) 3. Fronds narrower, up to 10 cm. broad 4 4. Fronds narrowly linear, distinctly stipitate, smooth, entire 5 4. Fronds irregular, more or less broadly expanded, obscurely stipitate, with undulate or ruffled margins 6 5. Fronds epiphytic, base broadly cuneate, apex not laciniate..l. P. hesperia (p. 517) 5. Fronds saxicolous, long stipitate, base narrowly cuneate, the apex deeply laciniate 2. P. fissilis (p. 518) 6. Fronds with undulate, much ruffled margins. .5. P. occidentalis (p. 520) 6. Fronds with smooth or slightly undulate margins... 3. P. latifolia (p. 519) 1. Punctaria hesperia S. and G. Plate 37, fig. 30, and plate 19, fig. 18 Fronds 1.5-2.5 cm. high, 5-10 mm. wide, solitary or in small clusters, tapering at the base to a very short stipe and to a blunt apex, 35-50/* (up to 80jLt) thick, composed of 4-6 layers of cells, those of the two middle layers much larger than those of the cortical layer ; cortical cells thin walled, 4-5 sided in surface view, 18-22/* diam. containing numerous parietal disk-shaped or small band-shaped chromatophores, 518 University of California Publications in Botany [Vol.8 medullary cells thin walled, closely joined, the length and breadth dimensions greater than the cross-dimensions, containing a few disk- shaped chromatophores ; zoosporangia numerous, promiscuously scat- tered among the gametangia on both surfaces of the frond; game- tangia very numerous, closely congested at times or scattered without order ; projecting hairs very delicate, in small groups, early deciduous. Growing on Phyllospadix. Pacific Grove, California. Setchell and Gardner, Phyc. Cont., 1924, p. 3. This species is apparently not abundant, however the plants are small and inconspicuous, and quite probably ephemeral. Of all of our species, thus far discovered, this one fruits most abundantly. The larger part of the surface cells become transformed into reproductive cells. The type, as stated above, comes from Pacific Grove, but we have two other collections from our territory, viz., growing on Phyl- lospadix at Victoria (Gardner, no. 38496) and one from San Pedro (Mrs. H. D. Johnston, Herb. Univ. Calif., no. 99456) on Zostera which may be referred here, at least provisionally. 2. Punctaria fissilis S. and G. Fronds fragile and flabby, ligulate, 12-15 cm. high. 10-16 mm. broad, 65-70/i. thick, tapering at the base to a very delicate distinct stipe 10-14 mm. long, attached by a small parenchymatous disk, light brown in color, more or less deeply lacerated into linear laciniae, com- posed of four layers of cells of fairly uniform size arranged in more or less longitudinal rows, with thin walls and very small disk-shaped chromatophores arranged closely along the walls; zoosporangia and gametangia unknown ; hairs present in small groups. Port Clarence, Alaska. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 4. The flaccid nature of the fronds, the long, delicate stipe, their laciniate habit, the arrangement of the cortical cells in fairly regular longitudinal rows along with the other combination of characters, characterize this as a distinct species. P. fissilis seems to have its nearest relations with P. glaciate Rosenv. and P. plantaginea (Roth) Grev. We have not been able to examine any of the material from Yakutat Bay collected by Saunders and reported under P. plantaginea (1901, p. 420). We have not seen any typical P. plantaginea from our territory and suspect that possibly his P. plantaginea may be found to belong to P. fissilis. 192 ->] Setehell-Gardner: Melanophyceae 519 3. Punctaria latifolia Grev. Fronds gregarious, flaccid, adhering firmly to paper, 1-3 dm. high, 2-7.5 cm. wide, 70-85//. thick, linear to obovate, attenuated at the base ; color yellowish to light brown; composed of four layers of cells, the outer layers smaller than the inner ; gametangia projecting but slightly above the surface ; zoosporangia, unknown. Occasional in quiet sandy coves (fide Saunders). Greville, Alg. Brit., 1830, p. 52 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 240. Homoeo stroma latifolium Saunders, Alg. Harri- man Exp., 1901, p. 420. We have examined a few small specimens of a Punctaria collected by Saunders at Metlacatla and on Baranoff Island, Alaska, and pub- lished (Joe. cit.) as P. latifolia, which seem to belong to this species. We are referring them here with some doubt on account of their much reduced size as compared with Wyatt's no. 9, of the Algae Dan- monienses, which we are taking a.s fairly typical of the species. The measurements of the cross-section of the fronds and the shapes and sizes of the cells composing the four layers agree very well with those of her distribution, both being 70/* to 80/j. in diameter. 4. Punctaria lobata (Saunders) S. and G. Fronds broadly linear, lanceolate or ovate, 10 or more cm. long, 1-5 cm. wide, narrowed below to a short stipe of a dark olive brown color, composed of 4 layers of cells, the 2 inner layers larger than the outer layers; edges deeply and irregularly lobed, the lobes irregularly cut and divided ; zoosporangia scattered over the whole surface of the plant ; gametangia unknown. Attached to Zostera marina. Prince William Sound and Sitka, Alaska. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 12. Homoeostroma looatum Saunders, Alg. Harriman Exp., 1901, p. 420. We have not seen any material belonging to this species and are including it here without comment, having previously made the above combination (loc. cit.). 520 University of California Publications in Botany t VoL - 8 5. Punctaria occidentalis S. and G. Plate 35, fig. 6, and plate 80b Fronds moderately rigid, linear-lanceolate or oblanceolate to broadly ellipsoidal, usually tapering at the base to a short stipe about 5 mm. long, attached by a small disk, 1.5-2 dm. high, 3-10 cm. broad, 40-180/a thick, with undulate and coarser ruffled margins, of a light brown color, soon changing to green on exposure, composed of 2-7 layers of cells, thickest in the median region, gradually thinner toward the margins ; cells of medulla largest diminishing slightly toward the surfaces, cortical cells being the smallest ; zoosporangia cuboidal to subspherical, scattered on both sides of the frond, 30-40//. (up to 70//,) diam., very slightly bulging on the outer surface; gametangia numerous, scattered promiscuously on both sides of the frond among the zoosporangia and usually protruding beyond the surface about one-half their length ; hairs unknown. Growing on Zostera, in the upper sublittoral belt. Monterey, California. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 4. Homoeostrama latifolia Saunders, Phyc. Mem., 1898, p. 159, pi. 30, figs. 4, 5. Punc- taria latifolia Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc.), no. 873; Tilden, Amer. Alg. (Exsicc), no. 355. We have not been able to observe the young fronds of this species, but the fruiting fronds have no hairs so far as we have been able to ascertain. Rosenvinge (1910, p. 121) has found a similar condition in his Punctaria glacialis. The surface is moderately sticky and the plants adhere firmly to paper on drying. As cited above, this species has been referred to P. latifolia since its discovery on our coast some years ago. We have not been able to examine any of the material upon which Greville founded the species. We have, however, examined a specimen distributed by Wyatt (no. 9) from Torbay on the southeast coast of England near the type locality. Since no other species have been cited from this general locality, and Wyatt 's specimen agrees so well with Greville 's original description, we are assuming this to be typical P. latifolia and using it as a basis of comparison. Our plants are much larger and thicker and have a larger number of layers of cells all of which are much more variable in shape than those of Wyatt 's distribution. The cortical cells which do not produce zoosporangia and gametangia usually divide hori- 1925] Setchell-Gardner : Melan&phyceae 521 zontally into two. The fructiferous cells remain undivided. The undivided cortical cells are about the same shape and size as the medullary cells which also vary in shape and size. 6. Punctaria chartacea S. and G. Plate 45, fig. 79, and plate 81 Fronds more or less finely bullate, membranaceous, chartaceous, linear to broadty elliptical, base often cordate, margins finely fimbriate, 25-35 cm. (up to 50 cm.) long, 8-12 cm. (up to 30 cm.) broad, mature plants 90-1 20ju, thick, attached by a very small disk ; composed of 6-7 layers of cells, the surface cells being much smaller than the medullary cells; cell Avails thick, of light brown color, with intercellular spaces among the medullary cells ; color dark brown, not changing on drying, not adhering at all to paper; hairs, zoosporangia and gametangia unknown. Growing on Eel grass in the lower littoral and upper sublittoral belts. Sitka. Alaska. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 4. This species of Punctaria was found in great abundance, both attached and floating. A large number of specimens were examined in the last part of June and no reproductive cells could be found. Neither were any hairs observed. 7. Punctaria expansa S. and G. Plate 49, fig. 19 Fronds gregarious, broadly expanded, more or less circular in out- line, 20-30 cm. broad, 100-1 30/x thick, very fragile, of medium brown color, fading to green very soon after removal from water, composed for the most part of four layers of cells, the medullary cells being much larger than the cortical cells; neither the medullary nor the cortical cells arranged in any definite series; cortical cells four to six sided and 30-40^ diam. in surface view ; zoosporangia and gametangia on the same frond, the former projecting but slightly beyond the sur- face, the latter more or less conical and projecting one-half or more of their length beyond the surface ; hairs present but sparse. Growing in a small quiet cove at the head of Penn's Cove, Whidbey [gland, Washington. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 5. Punctaria latifolia Setchell and Gardner, Alg. N.W. Amer., 1903, p. 240 (in part). 522 University o'f California Publications in Botany [Vol.8 Punctaria expansa is apparently a very rare though distinct species, the foregoing material referred to being the only collection known. The habit of growth is the same as that of a group of fair sized Vivas. It was found growing in a secluded cove in which the water rises and falls but slightly through a small channel with the change of tide, and is not disturbed by the action of wind. The plants are exceedingly fragile, being scarcely tough enough to hold their own weight when first removed from the water. They give up their brown color soon after being removed from the water when they superficially resemble almost perfectly an expanded Viva. The fruit, especially the zoo- sporangia in our specimens, is decidedly sparse, likewise the hairs. We have been able to examine a few small young specimens of Homoeostroma undulaUim collected and thus labeled by Saunders (1901, p. 419). They are not fruiting, are densely covered with diatoms and do not soak out readily. We are, therefore, on account of the state of the material, unable to place it and hence are not including it in our list of species. family 8. ASPEKOCOCCACEAE foslie Fronds ligulate or saccate, simple, differentiated into two kinds of tissue, an inner one, composed of nearly colorless cells and an outer cortical color bearing one; growth trichothallic at first, later inter- calary ; hairs present, and in some species, paraphyses ; zoosporangia on full-sized, or macroscopic, plants; gametangia, so far as known, on reduced, but not microscopic plants, not so definitely collected into sori as is the case with the zoosporangia ; both zoosporangia and game- tangia projecting from the surface and collected into sori scattered all over the frond except at the base. Foslie, Mar. Alg. Norway, Cont. I, 1890, p. 88. Asperococceae Zanard., Sagg. Class. Nat. Ficee, 1843, p. 10 (in part) ; Farlow, Mar. Alg. New England., 1881, p. 88. The Asperococcaceae, as limited to those ligulate, bullose, or nearly spherical forms, solid or hollow, with the zoosporangia in distinct and elevated sori, projecting and accompanied, usually at least, by distinct paraphyses or hairs and in having the gametangia, so far as known, in usually less definite sori and on plants more or less reduced in size, but still macroscopic, is closely related in its early trichothallic, later intercalary growth to the Striariaceae, Punctariaceae, and Scyto- siphonaeeae, so that it is occasionally difficult to determine exactly 1925] Setch ell-Gardner: Melanophyceae 523 where to place a particular genus. The genus Asperoeoccus, as exemplified by A. buUosus Lamour. is distinct and definite as is also Sormthera ulvoidea Kupr., and both of these are typical as to hollow frond, manner of growth and type of sorus which is definite in outline, with paraphyses and zoosporangia intermingled and with a group of hairs more or less centrally located. The genus Halorhipis of Saunders has also projecting sori with zoosporangia, but H. Wmstonii (C. L. Anderson) Saunders has less definite sori and is without paraphyses but provided with hairs. The California!! species seems to belong in Asperococcaeeae because of its superficial and projecting zoosporangia, and because of its general structure and appearance ought to be placed near to Asperoeoccus compressus Griff. We have seen no gametangia on the plants of any of our species, but judging from the statements of Buff ham (1891), Sauvageau (1895), and Kuckuck (1899), the gametophyte may be very much reduced, a tendency toward, but by no means approximating, the microscopic gametophyte known to occur in some of the other Melanophyceae. Key to the Genera 1. Fronds ligulate, solid 18. Halorhipis (p. 523) 1. Fronds cylindrical or more or less spherical, hollow 2 2. Fronds cylindrical or slightly clavate, more or less twisted 20. Myelophycus (p. 527) 2. Fronds spherical, sessile 19. Soranthera (p. 525) 18. Halorhipis Saunders Fronds solid, ligulate, arising from a disk-shaped holdfast, differen- tiated into two fairly distinct tissues, a medulla of several layers of larger, nearly colorless, closely united, more or less cylindrical cells extending lengthwise of the frond, and surrounding this a cortical or surface tissue composed of more or less cuboidal, color bearing cells ; zoosporangia superficial among hairs and collected into minute sori scattered abundantly over both sides of the frond except toward the base ; gametangia unknown. Saunders, Phyc. Mem., 1898, p. 160. Halorhipis was proposed by Saunders to receive a single species found at Pebble Beach, Monterey County, California. The genus is closely related to Haloglossum Kuetz. The genus Haloglossum was founded by Kuetzing (1843, p. 340) based upon the material of Asperoeoccus compressus Griffiths (in Hooker, 1833, p. 278) from the southern shores of England. He 524 University of California Publications in Botany [Vol. 8 changed the name to Haloglossum Griffithsianum which accordingly becomes the type of the genus. The chief distinction between Halo- glossum and Asperococcus lies in the character of the interior of the frond. The former is ligulate and solid and the latter is saccate. Mrs. Griffiths expressly states {loo. cit.) "there is not the slightest tendency to be tubular or inflated," in speaking of A. compressus. We feel inclined to keep the two genera separate, based mainly upon this character, although the method of reproduction in the two is similar. The genus Halorhipis is very similar in appearance and structure to Haloglossum. but the sori are destitute of paraphyses although the protruding zoosporangia are aggregated about a cluster of hairs. Halorhipis Winstonii (C. L. Anderson) Saunders Plate 35, fig. 8 Fronds mostly aggregated into tufts, thin, membranaceous, lanceo- late to obovate or spatulate, 8-20 cm. high, 2-5 cm. wide, about 200/x thick, with very slender filamentous stipes; color dark olive green; zoosporangia collected into sori mostly in slight depressions, ellipsoidal, obovate or pyriform, 30-45/x long, 20-30/x broad; sori numerous, variable in shape and size, linear to irregular in outline ; hairs in groups in the sori ; gametangia and paraphyses unknown. Growing on the fronds of Egregia Menziesm and on rocks. Known only from a single locality, Pebble Beach, Carmel Bay, California. Saunders, Phyc. Mem., 1898, p. 161, pi. 28; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 532. Punctarm Winstonii C. L. Anderson, Some new and some old algae, 1894, pp. 358, 359. Asperococcus Winstonii Svedelius, in Engler and Prantl, Natiirl. Pflanzenfam., Nachtr. z. 1 Th., Abt. 2, 1910, p. 158 (by implication). This plant, although known for only a little over a quarter of a century, has had a somewhat varied history. It was first described by C. L. Anderson (loc. cit.) who saw in it affinities with the genus Punciaria Grev. Saunders (loc. cit.) next studied it and came to the conclusion that a new genus, Halorhipis, should be erected for its reception. It was distributed in the same year under that name in the Phycotheca Boreali-Americana as no. 532. In 1910, as above stated, Svedelius placed it in the genus Asperococcus Lamour. Our studies, while leading us to retain the genus Haloglossum Kuetzing, discarded by most of the later phycologists, and to place Asperococcus compressus Griff, in that genus as a species distinct from all others, I 925 ] Setchell-Gardner: Melanophyceae 525 yet also lead us to keep llalorhipis Winstonii distinct because of the lack of the pluricellular paraphyses so characteristic of both Aspero- cocais and Haloglossum. It differs from a Punctaria, as we under- stand that genus, in having the thallus decidedly differentiated into two sorts of tissues, and in having the zoosporangia superficial. 19. Soranthera Post, and Rupr. Fronds spherical and solid in the juvenile stage, later becoming hollow, inflated and membranaceous, sessile, attached by a rather broad and penetrating base, composed of two kinds of tissue, an inner one 4-5 cells deep, of large, nearly colorless cells, and a surface layer of small, angular, color bearing cells; reproduction by zoo- sporangia in very distinct, conspicuous sori, scattered promiscuously over the surface of the frond ; paraphyses pluricellular ; hyaline hairs grouped in the center of the sori ; gametangia unknown. Postels and Ruprecht, Illustr. Alg., 1840, p. 19. The type species of the genus is S. ulvoidea discovered "Ad insulam Sitcha, parasitica in fronde Rhodomelae Laricis, " (fide Postels and Ruprecht, loc. cit.). The question of parasitism of the genus has been investigated by Miss E. S. Barton and was discussed by her in an article in the Journal of the Linnaean Society (1898, p. 479). She questions the statement quoted above from Postels and Ruprecht as to the parasitism of the species, taken in the sense that we understand parasitism. However, her investigation shows that the fronds are attached by a mass of rhizoidal filaments which sur- round the host, many of which actually penetrate into the cells even far into the interior of the host. She saw and figured what she inter- preted as ' ' haustoria. ' ' Whether these plants and many others among all four of the large sections of the algae which penetrate in a similar way, actually obtain food from the host has not been satisfactorily determined, so far as we are aware. Soranthera ulvoidea Post, and Rupr. Fronds membranaceous, comparatively thin, oval, ellipsoidal or nearly spherical, entire or irregularly lobed, up to 7 cm. diam. ; color olive brown; sori abundant, conspicuous, distributed fairly evenly over the whole free surface of the frond ; zoosporangia clavate, 70-100ju long, interspersed among numerous clavate, pluricellular paraphyses which surpass them in length. 526 University of California Publications in Botany [Vol. 8 Growing on species of Rhodomela and OdonthaUa, frequently in tide pools, in the littoral belt. Common along the whole coast from Unalaska, Alaska, to southern California. Postels and Ruprecht, Illus. Alg., 1840, p. 19; Kjellman, Om Beringh. Algfl., 1889, p. 47, pi. 7, figs. 4, 5; Saunders, Phyc. Mem. 1898, p. 165, pi. 29, figs. 4, 5; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 244. Two forms have been segregated as follows : Soranthera ulvoidea f. typica S. and G. Plate 39, figs. 40, 41, and plate 836 Fronds nearly regular in outline, membrane thicker, oval to spherical, up to 7 cm. diam. Growing on species of Rhodomela and Odonthalia. From Pacific Grove, California, northward into Bering Sea, and more common in the southern part of the range than f. difformis. Setchell and Gardner, loe. cit.; Kjellman, loc. cit.; Saunders, loc. cit., Alg. Harriman Exp., 1901, p. 422; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 417; Tilden, Amer. Alg. (Exsicc), no. 245. Soranthera ulvoidea f. difformis S. and G. Fronds gregarious, membrane thinner, usually smaller than the above, variously and deeply lobed. Growing on the same hosts as the two preceding, but is more common on the northern part of the range, from Cape Flattery northward. Setchell and Gardner, loc. cit.; Collins, Holden and Setchell, loc. cit., no. 1130. Forma difformis seems to prefer either brackish or muddy water or is induced by such medium. Soranthera ulvoidea Post, and Rupr. is a very common species along the entire Pacific coast of North America from Pacific Grove, California, northward. Careful search has failed to reveal the gametangia. 1925 J Setchellr-Gardner : Melan-ophyceae 527 20. Myelophycus Kjellm. Fronds filiform, attached by numerous rhizoidal filaments densely interwoven, solid throughout when young, later becoming hollow above; growth trichothallic (?), later usually intercalary; fronds composed of three kinds of tissue, an inner, colorless layer of a few rows of large, rounded, isodiametric, or elongated cells, outside of which is a layer of small prismatic, thick-walled cells giving rise to a dense mass of erect, assimilating filaments (paraphyses ?) ; zoo- sporangia immersed among the assimilating filaments; gametangia unknown ; hairs present in the juvenile stage. Kjellman, in Engler and Prantl, Natiirl. Pflanzenfam., 1893, 86 Lief., p. 202, Om Fuc. Myelophycus, 1893, pp. 1-11. The type species of the genus is M. caespitosum from the coast of Japan. The species of our coast seems to agree in all generic require- ments. Kjellman has followed the neuter gender of the Greek noun, phykos, although the Latin form would indicate a grammatical masculine. It seems best, however, to follow Kjellman as being more in accord with the proper usage. We place this genus in Asperococcaceae with considerable doubt. The species may prove to be the zoosporangial forms of gametangial plants similar to Chordaria Gunjii Yendo. Myelophycus intestinale Saunders Fronds solitary or loosely caespitose, 1-15 dm. high, 2-4 mm. diam., solid in the juvenile stage, hollow at maturity above the short, solid stipe, cylindrical, abruptly attenuated at the tip, but later eroded, usually twisted, unbranched, dark reddish brown in color; central tissue composed of 2-3 layers of elongated, cylindrical, colorless cells merging toward the outside into 3-4 irregularly arranged rows of thick-walled, cuboidal cells, these in turn giving rise to numerous paraphyses composed of 4-8 thick-walled cells and to zoosporangia ; zoosporangia ellipsoidal to obovate scattered over the frond except the stipe, 45-60/x long, 20-30/* broad. Growing attached to rocks in the lower littoral and upper sub- littoral belts. Shumagin Islands, Alaska, to Puget Sound, Washington. "Myelophycus imtestinalis" Saunders, Alg. Harriman Exp., 1901, p. 420 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 241. 528 University of California Publications in Botany [Vol. 8 Myelophycus intestinale presents a different aspect when young from that of the older plant. This is seen particularly at the tips which are long acuminate when young, but blunter and often worn away in age. The younger plants seem less conspicuously twisted than the older. Our plant has fewer layers of cells in the intermediate layer and shorter cortical filaments than the M . caespitosum of Japan. Myelophycus intestinale f. tenue S. and G. Plate 40, fig. 50 Fronds densely caespitose, inconspicuously twisted, 1.5-2.5 cm. high, 0.25-0.75 mm. diam. ; zoosporangia broadly ellipsoidal, 40-45/* long, 30-35ju broad. Growing on rocks, usually in shaded localities, and where the spray dashes against the rocks, along high-tide level. Coos Bay, Oregon, to central California. "Myelophycus intestinalis f. tenuis" Setchell and Gardner, in Gardner, New Pac. Coast Mar. Alg. I, 1917, p. 385. family 9. STEIAEIACEAE kjellm. Fronds filiform, solid, more or less branched, trichothallic, mono- siphonous at first, later usually of two sets of tissues, the inner of larger, colorless, more or less elongated cells, the outer a single layer of medium sized cells placed in horizontal rows and having chromato- phores; unilocular zoosporangia and plurilocular gametangia super- ficial or projecting beyond the surface in sori of more or less definite circumscription and usually in transverse rows; with or without paraphyses. Kjellman, Handbok I, 1890, p. 53. It seems best to follow Kjellman and keep the seemingly very distinct genera, Stictyosiphon, Striaria and their near relatives separate from the Asperococcaceae and the Scytosiphonaceae and retain them among the Striariaceae. Both the gametophytes and the sporophytes are known and are similarly macroscopic, but conditions are otherwise in the other two families just mentioned. In the Asperococcaceae there are dissimilarities in size, at least in some species, between the individuals bearing gametangia and those bear- ing zoosporangia, although neither is properly microscopic so far as known. Among the Scytosiphonaceae, only the gametangial form is known, suggesting that the zoosporangial form may be heteromorphic, 1925] SetcheU-Garrfncr: Melanophyceae 529 whether microscopic or eliminated by some method of syncopation may not even be surmised. Two genera have been reported from the western coast of North America, but there is so much doubt about one of them (Striaria, as indicated below) that we have taken only one under serious consideration. 21. Striaria Grev. Greville, Algae Brit. Syn., 1830, p. XLIIT. Striaria attenuata (Ag.) Grev. Greville (loe. tit.). Solenia attenuata Agardh., Syst., 1824, p. 187. A hand lens examination of the single scanty specimen from Orcas Island, Washington, in the Herbarium of Trinity College, Dublin, referred to Striaria attenuata, by Harvey (1862, p. 160) left the impression that it may belong rather under Stictyosiphon tortilis than under Striaria. It was very slender and seemed to be sterile. 22. Stictyosiphon Kuetz. Fronds attached by numerous branched rhizoids, mostly growing in dense clusters, filiform, branched, terminating in one or more hyaline hairs, solid in the juvenile stage, flstulose below, growth inter- calary ; plant body consisting of parenchymatous cells, the interior of more or less elongated, rectangular, colorless cells and the exterior of a single cortical layer of color bearing cells ; reproduction by unilocular zoosporangia and plurilocular gametangia produced by the trans- formation of certain cortical cells, solitary or more or less aggregated, borne on similar macroscopic plants. Kuetzing, Phyc. Gen., 1843, p. 301. The type species of the genus is S. adriaticux with the type locality Trieste. The northern species added later seem to belong under the same genus so that there is no necessity for retaining the genus Phloeospora of Areschoug (1873, p. 164). The species of Stictyosiphon differ principally from those of Striaria in not having paraphyses intermingled with the reproductive organs in their sori. Stictyosiphon tortilis (Rupr.) Reinke Fronds filiform, long attenuate, closely aggregated, repeatedly and alternately branched, the branches bearing minute ramuli mostly opposite, 5-8 cm. high, often with a small, longitudinal, central cavity in the lower portions; cortical cells with branched band-shaped chro- 530 University of California Publications in Botany [Vol.8 matophores; color dark olive brown; zoosporangia and gametangia scattered or aggregated into more or less transversely arranged groups. Port Clarence, Golofin Bay and Unalaska Bay, Alaska. Reinke, Algenfl. westl. Ostsee, 1889a, p. 55, Atlas, 1889, p. 47, pis. 31, 32; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 987. Phloeospora tortilis Setchell and Gardner, Alg. N.W. Amer., 1903, p. 245. Scytosipkon tortilis Ruprecht, Tange, 1851, p. 373. The plants of Bering Sea appear to belong in the form cycle of this Arctic species. We seem to remember having seen specimens from farther south along our coast, but cannot definitely place them. family 10. SCYTOSIPHONACEAE foslie Fronds cylindrical and hollow, flattened and solid, or more or less globular and hollow, main growth in length at first trichothallic, later intercalary, simple, of two or three tissues, one of internal hyphae in Endaraehne, but of large colorless cells in others, second, an inter- mediate layer of large colorless cells, surrounded by an outer layer of small, mostly cuboidal, colored cells in longer or shorter anticlinal rows ; zoosporangia unknown ; gametangia in extended and confluent sori of indefinite outline, palisade-like, of equal length and closely packed together, not projecting beyond the surface ; hairs in groups ; unicellular paraphyses ( ?) present in some species. Foslie, List Mar. Alg. Isle of AVight, 1892, p. 13. Scytosiphoneae Thuret, in Le Jolis, Liste des alg. mar. de Cherb., 1863, pp. 14 and 67. Thuret seems to have been the first to have stabilized the genus Seytosiphon and to have recognized it as the type of a group cor- responding to a family. Foslie, so far as we may learn, first used the family ending as now usually adopted. The Scytosiphonaceae are close to the several preceding families as has been noted previously. The method of growth is very similar, but only the gametangia are known in the genera attributed to this family and they are in expanded and indefinite sori which do not project above the surface of the frond nor are they associated with pluri- cellular paraphyses. The Punctariaceae possess the simplest members of this series and they have both zoosporangia and gametangia, these commonly produced on the same individual. In the Asperococcaceae, gametangial plants are rarely found and are smaller than the zoo- sporangial plants. In the Striariaceae both zoosporangial and game- tangial plants are found. While in the Scytosiphonaceae, although most of the species are common and widespread, no zoosporangial plants have been detected. I 925 ] Setchell-Gardner : Melanophyceae 531 Key to the Genera 1. Fronds solid and strap-shaped 2 1. Fronds hollow and cylindrical, flattened, globose or difform 3 2. Interior composed of hyphal tissue 25. Endarachne (p. 538) 2. Interior composed of parenchymatous tissue 24. Ilea (p. 535) 3. Fronds cylindrical or flattened 23. Scytosiphon (p. 531) 3. Fronds globose or difform 4 4. Membrane entire 26. Colpomenia (p. 539) 4. Membrane reticulately perforated 27. Hydroclathrus (p. 542) 23. Scytosiphon Ag. (emend. Thuret) Frond unbranched, solid when young, later tubular, cylindrical or compressed, constricted at short intervals or smooth throughout, con- sisting of two layers of cells, an inner of thick-walled, vertically elongated, colorless cells, and an outer of small, rounded, cuboidal, assimilating cells, the latter giving rise to the plurilocular gametangia covering the entire surface, except at the base, and to scattered uni- cellular paraphyses (?) and hairs when present; growth intercalary near the base. Agardh, Dispos. Alg. Suec, 1811, p. 24 (fide Pfeiffer) (pro parte) ; Thuret, Rech. sur les zoospores, 1850, p. 239. The genus Scytosiphon originally included what we now call Dictyosiphon foenicidaceus, Chorda Filum and Scytosiphon Lomen- taria. Thuret seems to have been the first to restrict the generic name to the Chord aria Lomentaria Lyngbye and to point out the generic characteristics. 0. Kuntze (1898, pp. 430 and 434) has proposed the generic name Tubicutis but the International Rules (Briquet, 1912, p. 76) decided in favor of the retention of Scytosiphon. Scytosiphon Lomentaria (Lyngb.) J. Ag. Plate 44, figs. 72, 74 Fronds usually fasciculate, sometimes solitarv and scattered, 15-30 cm. (up to 60 cm.) high, 3-6 mm. diam., cylindrical, at intervals more or less deeply constricted, attenuated below into a small solid stipe, attached by a small disk ; color dark brown to olive green ; plurilocular gametangia forming a compact palisade layer over the surface of the frond; paraphyses (?) scattered among the gametangia. Growing on rocks in the whole of the littoral belt. Common along the entire Pacific coast from Port Clarence, Alaska, to southern California. 532 University of California Publications in Botany [Vol. 8 J. G. Agardh, Sp. Alg., vol. 1, 1848, p. 126 ; Thuret, Rech. sur les zoospores, 1850, p. 239 (1851, p. 30, Repr.), pi. 29, figs. 1-4; Setchell, Alg. Prib. Islands, 1899, p. 591 ; Saunders, Phyc. Mem., 1898, p. 163, pi. 31, figs. 8-10, Alg. Harriman Exp., 1901, p. 421; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 243; Okamura, Icon. Japan. Alg.. 1908, p. 144, pi. 30. Chorda Lo-mentaria Lyngbye, Hydrophyt. Dan., 1819, p. 74, pi. 18, e. Seytosiphon Lomentaria, if all the forms referred to it belong under it, is a widespread and variable species. It varies from a centimeter or two to twenty-five or more centimeters in height and from a milli- meter or two up to ten millimeters in thickness. It may vary from clearly cylindrical to decidedly flattened and it may vary from regu- larly or irregularly and pronouncedly constricted (especially in the larger plants) to those with no trace of constriction whatever. It may be covered almost uniformly with gametangia or, rarely, the gametangia may be segregated into distinct and limited sori. The gametangia may be longer or shorter, with variations from two up to twenty-five or more longitudinal uniseriate loculi. It has been the custom to separate the var. complanata Rosenvinge, but this occurs in a major and minor form, as do all the variants with the exception per- haps of the typical form (f. typicus) which is constricted, but while this varies much in size, it never occurs among the true minor forms of the species. The larger and smaller forms of each series of forms differ from one another in the size of the cells of the inner layer, and in the length (absolute and in number of loculi in each longitudinal series) as well as in diameter of the gametangia. In order to coordinate the variables as mentioned above, we have tabulated them in the fol- lowing key, rather to give a view of the combinations of characters than that they should be considered of any definite taxonomic import- ance, since intermediate forms are, by no means, absent. Key to the Forms 1. Plants constricted at intervals, medium to large 1. f. typicus (p. 533) 1. Plants not constricted 2 2. Plants cylindrical 3 2. Plants complanate 5 3. Plants 6 cm. or more high; interior cells large; gametangia long 2. f. cylindricus major (p. 533) 3. Plants less than 6 cm. high; interior cells small; gametangia short 4 4. Sori covering the frond 3. f. cylindricus minor (p. 533) 4. Sori restricted, definite 4. f. cylindricus maculatus (p. 534) 5. Plants large, 6 cm. or more; inner cells large; gametangia long 5. f. complanatus major (p. 534) 5. Plants small, less than 6 cm.; inner cells small; gametangia short 6. f. complanatus minor (p. 534) l925 ] Setohell-Oardner: Melanophyceae 533 The occurrence or non-occurrence of hairs is such a variable char- acter as is also the presence or absence of the ''unicellular paraphyses" (possibly hypertrophied gametangia ?) in all of our variants of the species that we have not found it possible to use either structure in any of them as a diagnostic character of much significance. 1. Scytosiphon Lomentaria f. typicus S. and G. Plate 39, fig. 45, and plate 44, fig. 75 Plants of the largest size of the genus, both as to length and diameter, dark brown to olive green in color, more or less regularly and deeply constricted; paraphyses ( ?) cylindrical to slightly clavate, not projecting beyond the gametangia, with chromatophores often collected at the outer end ; gametangia with 10-15 loculi. Growing on rocks in pools in the lower littoral belt. Port Clarence, Alaska, to southern California. Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 323&^ Tilden, Amer. Alg. (Exsicc), no. 246. 2. Scytosiphon Lomentaria f. cylindricus major Fronds from 6-25 cm. long. 1.5-2.5 mm. diam., not constricted, cylindrical, long-attenuated at both ends, light brown in color. Growing on rocks in the lower littoral or in tide pools in the upper littoral belts. Central California. Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1389. Chordaria attennata Tilden, Amer. Alg. (Exsicc), nos. 347a, 3476? The cylindrical unconstricted forms seem to be fairly well estab- lished both as regards habitat and form. They grow in tide pools, often even above high-tide level, and the size seems to be materially influenced by the size and depth of such pools. In these pools where they are frequently to be found, they are generally far removed from the typical form which grows usually in deep tide pools in the middle and lower littoral belts. 3. Scytosiphon Lomentaria f. cylindricus minor Fronds less than 6 cm. high, 1-1.5 mm. diam., cylindrical, uncon- stricted. Growing on rocks along high-tide line in localities exposed to wave action. Central California (San Francisco). 534 University of California Publications in Botany [Vol.8 4. Scytosiphon Lomentaria f. cylindricus maculatus Plate 42, figs. 61, 62 Fronds 3-4 cm. high, cylindrical, not constricted, dark brown in color ; gametangia collected into definite, larger or smaller sori. Growing on rocks exposed to heavy surf along high-tide level. Central California (San Francisco). Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1235. This form is distinguished by the presence of gametangia in groups, giving the small fronds a warty appearance. It grows in exposed situations in which the rocks are kept more or less moist continuously by the spray, but for the most part the plants are not submerged in pools. 5. Scytosiphon Lomentaria f . complanatus major Fronds flat, 9-14 cm. high, 1-2.5 mm. wide, dark brown in color, soon changing to green on removal from the water. Growing on rocks in the lower littoral belt. Pebble Beach, Carmel Bay, Monterey County, California. 6. Scytosiphon Lomentaria f . complanatus minor Plate 44, fig. 74 Fronds 3-5 cm. high, up to 1.5 mm. wide, irregular and variable in thickness, cylindrical below and solid or hollow, flattened above and hollow, light brown in color ; gametangia composed of 5-7 loculi, completely dissolving at maturity leaving the medulla clean ; medulla composed of two layers of cells, variable in size, much longer than broad ; hairs and paraphyses absent. Growing on stones along high-tide level or above. Fort Point, San Francisco, California. This form is a short-lived winter and spring form, usually matur- ing in February and March, so far as our experience shows. It has been detected for many years in the above mentioned habitat, often found growing on wood as well as on a stone wall. 1925 ] Setch ell-Gardner: Melanophyceae 535 24. Ilea Fries Fronds solid, or rarety with occasional small cavities, membran- aceous, considerably flaccid, simple, ligulate, tapering at the base to a small, solid stipe, attached by a parenchymatous disk-shaped holdfast, or by rhizoids, differentiated into two distinct tissues, a medulla com- posed of more than one layer of mostly large, more or less colorless cells and a cortex of small, cuboidal, color bearing cells ; the cortical cells on both surfaces giving rise by numerous horizontal divisions to gametangia with uniseriate loculi, gametes escaping by the complete dissolution of the entire wall; zoosporangia unknown; paraphyses( ?) and hairs present in some species. Fries, Flor. Scan., 1835, p. 321 (pro parte) ; not Ilea Fries, Syst. Orb. Veg., part 1 (pi. homon.), 1825, p. 336; not Ilea J. G. Agardh, Till Alg. Syst., afd. 3, 1883, p. 114. Phyllitis Kuetzing, Phyc. Gen., 1843, p. 342 (not Phyllitis Hill). Petalonia Derbes et Solier, Sur les org. repr. des alg., 1850, p. 265. As to the confusion and the different views concerning the adoption of Ilea as the name for this genus, one may consult Nordstedt (1911, p. 265) and M. A. Howe (1914, p. 51). Petolonm is preferable from the point of view of possible confusion, but Ilea seems to have the right of way. The genus tends toward complanate forms of Scyto- siphon in certain forms which, while not completely hollow, never- theless are not completely solid. The type species is Ilea Fascia (Muell.) Fries, whose type specimens come from somewhere on the coast of Norway. In the genus Ilea, a condition exists very similar to that found in Seytosiphon and in Colpomenia, that is, of a number of fairly striking form-types between which it is extremely difficult to make other than very arbitrary distinctions. We are inclined to recognize a single species, Ilea, Fascia, and refer the other described species or forms to it. Ilea Fascia (Muell.) Fries Plate 44, figs. 68-71, 73 Fronds attached by a small, parenchymatous disk, often fasciculate and several disks coalescing, thin, plane or at times more or less crisped, linear-lanceolate to oblanceolate, often rounded and more or less eroded above, 6-12 cm. (up to 25 cm.) high, mostly 1.5-2.5 cm. wide, but at times up to 12 cm. wide, exceedingly variable in thickness; 536 University of California Publications in Botany [Vol. 8 gametes escaping by the simultaneous dissolution of the entire cell walls of the gametangia. Growing on rocks and on Phyllospadix in the littoral belt. Occur- ring in its various forms from Unalaska, Alaska, to southern California. Fries, Flor. Scan., 1835, p. 321. Phyllitis Fascia Kuetzing, Phyc. Gen., 1843, p. 142, pi. 24, III, figs. 1-6, in Linnaea, vol. 17, 1843, p. 97 ; Saunders, Alg. Harriman Exp., 1901, p. 421 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 243; Farlow, Anderson and Eaton, Alg. Exsicc. Amer.-Bor., no. 199. Pctalonia Fascia. Kuntze, Rev. Gen. PL, vol. 3, 1898, p. 419 ; Howe, Mar. Alg. Peru, 1914, p. 50. Fucus Fascia Mueller, in Flor. Dan., 1778, pi. 768. So far as we have been able to ascertain, typical Ilea Fascia, has neither hairs nor paraphyses. The Phyllitis Fascia of Okamura (1901, pi. 10) shows a group of hairs among plurilocular gametangia. Later, however, he distributed this plant under Endarachne Binghamiae, in his Algae Japonicae Exsiccatae, no. 86, referring to his previous publication (loc. cit.) as a synonym. The size of the plants, as noted by different authors, is exceedingly variable, as the foregoing diagnosis shows. We have included plants up to 25 cm. long, and up to 12 cm. wide. The thickness varies with age in the same plant, and varies in different collections of plants. Mueller's Fucus Fascia' (loc. cit.), as illustrated, is about 20 cm. long and 5 mm. broad. The f. zosterifolia in size, form, and in being inter- ruptedly and slightly hollow, approaches Seytosiphon Lomentaria f. complanatus of medium size. We suggest the arrangement of the forms of the species in the following kev : 'r- 1. Frond broad, stipe distinct 2 1. Frond narrow to almost filiform, almost stipeless 3 2. Tapering gradually at the base to a short stipe .... 3. f. caespitosa (p. 537) 2. Tapering abruptly at the base to a short stipe 2. f. debilis (p. 537) 3. Frond flat, solid throughout 4 3. Frond with few small internal spaces 4. f. zosterifolia (p. 537) 4. Frond flat, attached by a solid disk 1. f. typica (p. 537) 4. Frond nearly filiform, attached by rhizoids 5. f. filiformis (p. 537) With the exception of the last, the forms thus far described for the genus, whether as species or varieties, seem to be represented in our territory. 1925] Setchell-Gardner : Melanophyceae 537 1. Ilea Fascia f. typica S. and G. Setehell and Gardner, Phyc. Cont., VII, 1924, p. 12. f. typica is founded on Fucus Fascia Mueller, Fl. Dan., 1778, pi. 768. The plants distributed in Collins, Holden and Setehell, Phyc. Bor.-Amer. (Exsicc.), no. 1131, are fairly representative of the typical form. It extends along the coast from Puget Sound, Washington, to southern California. 2. Hea Fascia f. debilis S. and G. Setehell and Gardner, Phyc. Cont., VII, 1924, p. 12. f. debilis is founded on Laminaria debilis Agardh. We have a few specimens representative of this form from San Juan Island, Wash- ington. They measure up to twelve centimeters wide and twenty-five centimeters long*. *&■ 3. Ilea Fascia f. caespitosa S. and G Setehell and Gardner, Phyc. Cont., VII, 1924, p. 13. f . caespitosa is founded on Laminaria caespitosa J. G. Agardh, Sp. Alg., I, 1848, p. 130, and is well illustrated in Thuret and Bornet, Etudes Phyc, 1878, pi. 4. The plants distributed in Collins, Holden and Setehell, Phyc. Bor.-Amer. (Exsicc), no. 736 may be considered to represent this form. It is known only in the extreme northern portion of our region. 4. Ilea Fascia f. zosterifolia S. and G. Setehell and Gardner, Phyc Cont., VII, 1924, p. 13. f. zosterifolia is founded on Phyllitis zosterifolia Reinke, Algenfl. westl. Ostsee, 1889a, p. 61. This form has been detected in but a single locality on our coast, viz., Pebble Beach, Carmel Bay, Monterey County, California. 5. Ilea Fascia f. filiformis S. and G. Setehell and Gardner, Phyc Cont., VII, 1924, p. 13. f. filiformis is founded on Phyllitis filiformis Batters, in Linn. Soc Journ. Bot, vol. 24, 1888, p. 451, pi. 18, figs. 1-6. This form has not been detected on our coast. 538 University of California Publications in Botany [Vol. 8 25. Endarachne J. Ag. Fronds plane, entire, ecostate, solid, attached by a small solid disk, composed of three quite distinct tissues, a surface tissue of more or less cuboidal cells holding the plastids, merging below into larger, thick-walled, slightly parenchymatous cells which in turn merge into the medulla, composed of very thick-walled, densely intertwined, branched filaments extending longitudinally for the most part ; repro- duction by plurilocular gametangia forming a palisade layer over the whole frond except the stipe ; paraphyses unknown ; hairs in fascicles. J. G. Agardh, Analecta Alg., Cont. 3, 1896, p. 26. The type species of the genus is Endarachne Binghamiae collected by Mrs. Bingham in the vicinity of Santa Barbara, California. So far as is known at present, this is the only species of the genus. In the genus Ilea, the central cells of the medulla are, at times, more or less elongated, but there is no such conspicuous development of hyphal tissues as is to be found in the present genus. Endarachne resembles forms of Ilea Fascia so closely in general appearance that it is necessary to examine sections to distinguish them with certainty. Endarachne Binghamiae J. Ag. Plate 38, figs. 37, 38, and plate 83a Fronds usually aggregated into clusters, 10-18 cm. high, 1-3 cm. broad, linear to broadly spatulate, at times irregular in outline, usually eroded above at maturity, with cuneate base and distinct, small, short stipe ; color dark brown, young plants usually turning green on drying ; gametangia formed by horizontal division of surface cells, 44-50/* long, 4-5/x diam., with uniseriate loculi, all enclosed by a common cuticular layer. Growing on rocks in the middle and upper littoral belts. Southern California and as far as Ensenada, Lower California. J. G. Agardh, Analecta Alg., Cont. 3, 1896, p. 26, pi. 1, fig. 5; Saunders, Phyc. Mem., 1898, p. 162, pi. 30, figs. 6, 7 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1593; Okamura, Alg. Japon. Exsicc, no. 86. Phyllitis Fascia Okamura, Illus. Mar. Alg. Japan, vol. 1, 1901, pi. 10. 1925] S etch ell-Gardner: Melanoplwjceae 539 26. Colpomenia Derb. and Sol. Frond solid in the juvenile stage soon becoming hollow, rather thin and membranaceous, whole when young, becoming irregularly, more or less lacerated in age, consisting of two tissues, an inner tissue of large, rounded, nearly colorless, thin-walled cells and a cortical tissue of small, more or less cuboidal, assimilating cells; plurilocular game- tangia in early stage of development growing around groups of hair filaments, later spreading over the entire surface of the frond ; para- physes (?) unicellular, clavate, scattered. Derbes and Solier, Mem. phys. alg., 1856, p. 11. This is the third genus among the Scytosiphonaceae to show a mul- titude of variations without sharp distinctions and no less than five forms, varieties, or species have been described and named. The type of the genus is conceived of as nearly a globular plant, thin-walled and while, at times, slightly bullose, at least neither thick-walled nor pro- vided with conspicuously projecting, wart-like or finger-shaped lobes. Nevertheless, such variations occur, especially on our coast, where the variation within the genus seems to have reached its maximum. We have placed all of these species, varieties, and forms under one species, with a series of varieties, much as we have done in the case of Scyto- siphon and Ilea, but these variations, being more pronounced, have been segregated into varieties and forms rather than into forms and subforms. Nevertheless they may be simply environmental effects or "ecads." Colpomenia sinuosa (Roth) Derb. and Sol. Plate 45, figs. 82-86 Frond sessile, attached by a broad base, thin and membranaceous, 4-10 cm. diam., filled with water when young, later collapsed and somewhat complanate, spherical to irregular in form, 0.25-0.35 mm. thick ; color dark brown ; cortical layer of cells consisting of 1-2 rows of cuboidal or polygonal cells, the inner layer of 2-5 rows of larger, more or less rounded cells; plurilocular gametangia 18-22/i. long, 5-8^ broad, the loculi frequently in two rows. Growing on rocks and on other algae in the middle and lower littoral belts. Common along the whole Pacific coast, from Yakut at Bay, Alaska, to southern California. 540 University of California Publications in Botany [Vol. 8 Derbes and Solier, Mem. phys. alg., 1856, p. 11, pi. 22, figs. 18-20 ; Saunders, Phyc. Mem., 1898, p. 164, pi. 32, figs. 7, 8, Alg. Harriman Exp., 1901, p. 421; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 242; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 278; Tilden, Amer. Alg. (Exsicc), no. 522. Viva sinuosa Roth, Cat, Bot.. vol. 3, 1806, p. 327, pi. 12. The type of the species was collected at Cadiz by Mertens and from the figures and what we know of the plants of that region is evidently of the more regularly globular or lobed, thin-walled, more or less smooth form of the species, although the description of Roth might seem to apply to the more thick-walled and rugose form. The typical form is less common on our coast but is found in certain localities. As stated above, it seems best, after a study of all of our west coast material, to consider all of our forms under one species and we have arranged them according to the following key : 1. Fronds thin, membranaceous, surface nearly smooth 2 1. Fronds thick, chartaceous, surface bullose or strongly lobed 3 2. Fronds approximately globose, very thin and smooth 1. f. typica (p. 540) 2. Fronds elongated or irregular, somewhat thicker and less smooth 2. f. expansa (p. 540) 2. Fronds expanded to several decimeters, thin but with short, spine-like tubercles 3- f. expansissima (p. 541) 3. Fronds sinuose, strongly bullose and tuberculate, but never lobed 4. var. tuberculata (p. 541) 3. Fronds extending into one to several long, finger-like lobes 5. var. deformans (p. 542) 1. Colpomenia sinuosa f. typica nom. nov. The typical form is thin, about 25-30^ thick, fairly regular in form, usually attached at one place to a filamentous alga, and growing, most commonly, in fairly quiet waters. The clusters of hairs are not sunk in the tissue of the frond and the layers of colorless cells are few (usually 2 to 3). 2. Colpomenia sinuosa f. expansa Saunders Fronds aggregated forming expansions of indefinite size and shape, 0.35-0.45 mm. thick, inner layers of cells composed of 5-7 rows of cells. Near Avalon Bay, Santa Catalina Island, California. Saunders, Phyc. Mem., 1898, p. 164, pi. 32, figs. 4-6; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 825. 1925] SetcheU-Gardner: Melanophyceae 541 The f. expansa is a variation of the typical form in which the frond is more ample, less regular in shape, usually creeping over filamentous algae and forming secondary attachments. It has a smooth surface in the sense of not being tuberculate or conspicuously lobed. It is somewhat thicker than the typical form, with the clusters of hairs sunk in tissues. The layers of colorless cells are greater in number than in the typical form (usually 5 to 7). 3. Colpomenia sinuosa f. expansissima S. and G. Fronds 3-6 dm. diam., thin, sinuose, with minute, spine-like pro- jections. Floating in billowy masses in San Francisquito Bay, Lower Cali- fornia. Setchell and Gardner, Mar. Alg. Gulf of Calif., 1924, p. 726. This form is probably an extreme of f. expansa, but it has only been found floating. It is thinner and less even in outline and surface, but is similar in every other way. The only specimens are sterile and show no hairs. 4. Colpomenia sinuosa f. tuberculata (Saunders) S. and G. Fronds sessile, hollow, rigid and somewhat coriaceous, very irregu- lar in shape and size, usually somewhat flattened, 5-10 cm. diam., 1-2 mm. thick ; surface much convoluted, wrinkled and folded, at maturity covered w-ith blunt warts or tubercules ; color dark brown ; cortex of 3-5 rows of cuboidal cells, the inner layer of 5-8 rows of large, irregular, thin-walled cells; plurilocular gametangia 20-25/a long, 3-4/a wide ; paraphyses 22/a long, 5/a wide. Growing on rocks and on other algae, often aggregated in large, brain-like masses, in the middle of the littoral belt. From Unalaska, Alaska, to La Paz, Lower California. Setchell and Gardner, Alg. N.W. Amer., 1903, p. 242 ; Howe, Phyc. Stud. V, 1911, p. 495; Setchell and Gardner, Mar. Alg. Gulf of Calif., 1924, p. 725. Colpomenia tuberculata Saunders, Phyc. Mem., 1898, p. 164, pi. 32, figs. 1-3 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 826. The Colpomenia tuberculata Saunders is thick-walled, of general globose form, but w T ith the distribution of growth areas interrupted in such a way as to give rise to bullosities and tumor-like warts. Some- times these swellings develop and fall out, leaving holes or lacunae, a suggestion of the process in Hydrorfathrus elathratus. 542 University of California Publications in Botany L VoL - 8 5. Colpomenia sinuosa f. deformans S. and G. Fronds rigid, erect, membranaceous, cavernous, lower part tuber- culate in the juvenile state, later developing one or more irregularly cylindrical, sack-like, at times flattened, projections from the upper surface, 1-7 cm. high, surface of frond smooth, somewhat lacerated when aged ; color dark brown. Attached to rooks and to other algae by a broad sessile base, in the lower littoral belt. Ranging from Cook Inlet, Alaska, to the Gulf of California. Setchell and Gardner, Alg. N.W. Amer., 1903, p. 242, pi. 18, figs. 13-15 ; Mar. Alg. Gulf of Calif., 1924, p. 726, pi. 19, figs. 61, 62. Scyto- siphon bullosus Saunders, Phyc. Mem., 1898, p. 163, pi. 31, figs. 1-7, Alg. Harriman Exp., 1901, p. 421; Tilden, Amer. Alg. (Exsicc), no. 351. The Scytosiphon bullosus Saunders always arises from a sinuous, lobed base which may be so small as to seem almost like a disk, but varies to a Colpomenia sinuosa type of considerable size. We have already discussed this elsewhere (1903, p. 243, pi. 18, figs. 13-15). When the long finger-like lobes are reduced to one and the surface is smooth, it is the typical form of Saunders, but more often the long lobes are several and unequal and, at times, all are short. It seems that certain areas grow more rapidly and this differentiation completes the series from the forms of var. typica to the extremes of var. deformans. 27. Hydroclathrus Bory. Fronds spherical or irregularly ovate, hollow, similar to those of Colpomenia, entire when young, at first becoming fenestrate, but broken later into a well formed lattice work, the meshes being variable in shape and size ; reproduction by plurilocular gametangia scattered over the whole outer surface of the frond ; hairs growing in groups in shallow depressions over the outer surface of the frond. Bory, in Diet, class., vol. 8, 1825, p. 419. The type and only known species of the genus is Hydroclathrus clathratus. The completeness of the perforation of the thallus of the single species of the genus Hydroclathrus readily separates it from the various forms of Colpomenia, although odd plants of C. sinuosa var. tuberculoid often become considerably, but not regularly, perforated through spotwise localization of growth. The origin of the holes has 1925 J Setch ell-Gardner: Melanophyceae 543 not been investigated. When first visible, they are very small and it seems likely that they originate through the death, or quiescence, of a cell or small group of cells in the midst of a rapidly growing area, as is the case in some other perforated algae. Hydroclathrus clathratus (Bory) Howe Fronds very irregular in form, in age much lacerated, 5-16 cm. diam., sessile, attached to the substratum by a broad base, sometimes several bases confluent, differentiated into two layers of cells ; the apertures varying much in size and form and the frond involute along their margins ; color yellowish brown ; hairs in small tufts in the center of expanded sori. Growing in tide pools in the lower littoral belt. On the west coast of Mexico near the mouth of the Gulf of California. Howe, in Britton and Millspaugh, Bah. Flor., 1920, p. 590 ; Setchell and Gardner, Mar. Alg. Gulf of Calif., 1924, p. 727. Hydroclathrus cancellatus Bory, in Diet, class., vol. 8, 1825, p. 419; Harvey, Phyc. Austral.. 1859, Tab. 98 • Mitchell, in Murray, Phyc. Mem., part 2, 1893, pp. 53-56, pi. 15, figs. 2—4. Eneoelium clathratwm (Bory) Agardh, Sp. Alg.. vol. 1, part 2, 1822, p. 412. Fucus clathratus Bory, MS., fide Agardh, loc. cit. The single species of Hydroclathrus is tropical and widespread in both the Atlantic and the Pacific oceans. Our specimens, from the Gulf of California, are old and we have been unable to detect any gametangia. family 11. 2EGIRACEAE fam. nov. Fronds erect, more or less branched, from slightly to decidedly gelatinous; main growth in length trichothallic, with central axes monopodial and cortical axes sympodial, of two or three tissues, (1) medullary of larger and longer colorless longitudinal filaments, easily separable or compact, (2) intermediate of broad but short colorless cells, and (3) cortical of longer or shorter distinct anticlinal filaments, often curved and enlarged upwards ; both unilocular zoosporangia and plurilocular gametangia known (in some species at least) and borne on similar (macroscopic) plants. The family which we have felt constrained to designate as Mgira- ceae is to be separated both from the Chordariaceae (in the narrower sense of Chordaria as limited by the type C. flagelliformis) and from 544 University of California Publications in Botany [Vol.8 the "Mesogloiaceae" (also in the narrower sense of the genus Meso- gJoia, as limited by the type species of the genus M. vermiculata) both because of the method of terminal growth ("trichothallic" as con- trasted with "subapical") and because neither in Chordaria nor Mesagloia (as limited to conform with the type species in each case) are plurilocular gametangia known, while they are known (or reported) in the typical species of MyriocJadia and Liebmannia. There seems to be represented here a distinct group agreeing in character- istics of typical trichothallic apical growth and with very similar mac- roscopic gametophytes and sporophytes although more exact knowl- edge of the types of these genera is extremely desirable, since the various statements are perplexing and authentic material difficult (or impossible for us) to obtain. We have made, therefore, an arrange- ment of our genera and species based on the most certain information available to us. It is to be borne in mind that MesogJoia (based strictly on the type of the species) has been removed to the Chordariaceae where it is close to, if not identical with, the genus Chordaria. Other species than the type, referred by different authors to Mesogloia are referred variously to Liebmannia, Mgira, MyriocJadia, MyriogJoia and Chordaria. Key to the Genera 1. Gametangia arranged unilaterally on the upper portions of the cortical filaments 28. .ffigira (p. 544) 1. Gametangia not unilateral, arising from the lower portions of the cortical fila- ments 29. Meneghiniella (p. 548) 28. .flSgira Fries Fronds branched, filiform, with trichothallic growth, composed of a medulla of colorless, branched filaments firmly agglutinated, giving rise to short, cortical, assimilating filaments and to basal zoosporangia, or to gametangia unilaterally arranged on their outer extremities or, to both zoosporangia and gametangia. Fries, Syst. Orbis, 1825, p. 342. CJadosiphon Kuetzing. Phyc. Gen., 1843, p. 329. Castagnea Derbes and Solier, Sur les organes repr. des algues, 1850, p. 269. Eudesme, J. Agardh. Till Alg. Syst. II, 1880, p. 29. Kuetzing founded the genus Cladosiphon (1843. p. 329) on a Mediterranean species, C. mediterraneus, which seems definitely, as stated below, to be either the same or most certainly closely related to, the species on which Derbes and Solier founded their Castagnea. 1925] Setchell-Gardner : Melanophyeeae .14") J. G. Agardh (1882, p. 40) took up Cladosiphon from the point of view of Meneghini's Liebmannia Posidoniae, a species seemingly at present, at least, to belong to another genus (Meneghiniella) from the type of Kuetzing. Derbes and Solier founded the genus Castagnea on a species they named C. polyearpa, presumably from the neighborhood of Marseilles, but the locality is not mentioned. Tbis species has the gametangia produced as outgrowths from the convex sides of the upper cells of the curved cortical filaments and this peculiarity is made a part of the diagnosis of the genus. Derbes and Solier state that their ^enus Castagnea is readily distinguishable from the related genera Lieb- mannia, Stilophora, and Nereia by the form of the filaments associated with the organs of fructification in that they are cylindrical and not terminated by large swollen cells. Castagnea polyearpa Derbes and Solier has been variously and usually doubtfully referred by later writers and we have found no reference to the existence of an authentic specimen. Derbes and Solier compare its appearance to that of a stunted Liebmannia. In their later (?) paper, Derbes and Solier (Sur les organes repr. des algues, 1850, p. 269, pi. 33, figs. 12, 13) seem inclined to consider their C. polyearpa. closely related to, if not identical with, the "Cladosiphon fistulosum" of Kuetzing, a binomial unknown to us, but presumably the Mesogloia fistulosa Zanard. (in Meneghini, 1843, p. 292) and considered by Kuetzing (1849, p. 547) to be a synonym of his Cladosiphon mediterraneus. There is certainly suspicion here that Castagnea and Cladosiphon may have been founded on the same type species. In the absence of that exact proof which is difficult to obtain other than from type specimens and experience from type localities, we feel justified in considering these two genera as identical in foundation. We may call attention to the fact that Bornet (1892, p. 236) has placed the Cladosiphon mediierraneus Kuetzing under Castagnea. The genus Eudesme was founded by J. G. Agardh (1882, p. 29) on the Mesogloki vireseens Carmichael, i.e., at least technically. There is good reason to believe that Agardh did not restrict his ideas to type, but had a composite idea as to the nature of his E. vireseens. He speaks of an "axis solidescens" whereas Eudesme has been applied as typically including species whose axial filaments are readily separable from one another (cf. Kjellman, 1893, pp. 225, 226) and as they are, to a certain extent at least, in what may be considered to be typical E. vireseens, viz., Alg. Danm., no. 49; but even here there is a definite 546 University of California Publications in Botany L VoL - 8 pseudoparenchymatous axis (cf. also Harvey, Phyc. Brit., pi. 82, 1847). Agardh also emphasizes in his generic diagnosis the formation of gametangia as projecting ramelli arising from the upper cells of the cortical filaments. The illustrations of Thuret (1850, pi. 27) show exactly the type of cortical filaments and gametangial formation upon which, as distinctive characters, Derbes and Solier founded their genus. So far as we are able to discover, the three genera, Castagnea, Cladosiphon, and Eudesme are to be united and, since Cladosiphon is the earliest name, unless we adopt Mgwa Fries of doubtful status, we might feel compelled to use Cladosiphon as the name of the genus to which we may assign species usually referred to Castagnea or to Eudesme. Cladosiphon, as we understand Kuetzing's type species, CI, meditcrraneus {Mesogloia fistulosa Menegh., Castagnea fistulosa (Menegh.) Derb. and Sol., C. polycarpa Derb. and Sol. and Castagnea mediterranean (Kuetz.) Bornet) is however, usually understood in the sense of J. G. Agardh (1882, pp. 40 and 42). J. G. Agardh, on the other hand, based his idea of Cladosiphon on Liebmannia Posidoniae Meneghini (1843, pp. 300-304, pi. 5, fig. 1) which has its gametangia situated very differently from those of the members of the Clado- siphon-Castagnea-Eudesme complex. Meneghini 's species seems to have the gametangia arising as transformed branches of the cortical filaments and slender, with uniseriate loculi, not pronouncedly sili- quaeform as in Liebmannia and not arising unilaterally from the upper cells of the cortical filaments as in the complex we are discussing. We have one species, from the Gulf of California, seemingly cogeneric with Meneghini 's Liebmannia Posidoniae for which we have proposed the generic name Meneghiniella (Phyc. Cont. VII, 1924, p. 5). It remains to consider the oldest name possible for this Cladosiphon- Castagnea-Eudesme complex and that is .Egira Fries (1825, p. 342). Fries founded his genus on the Linckia Zosterae Lyngbye (1819, p. 194, pi. 66, fig. C) which may or may not be the Rivularia Zosterae Weber and Mohr (1810, p. 367). The plant of Lyngbye has been variously interpreted by J. G. Agardh (1848, p. 53) and Areschoug, who had an authentic specimen for comparison (Areschoug, Alg. Scand. Exsicc, ser. 1, no. 67, and 1842, pp. 228-230). The evidence is well discussed by Kylin (1907, pp. 85-87). While there may- be reasonable doubt as to whether Lyngbye 's plant is the same as Eudesme virescens (Carm.) J. Agardh, or Eudesme Zosterae (J. Ag.) Kylin, it is to be placed in the Cladosiphan-Castagnea-Eudesme complex which ought, therefore, to take the name JEgira (cf. also, M. A. Howe, 1918, p. 505). 1925] SetcheU-Gardner: Melanophyceae 547 Since the foregoing was written, we have been allowed, through the kindness of Professor C. H. Ostenfeld, to examine one of Lyngbye's specimens of Linckia Zosterae. It is clearly very closely related to Eudesme virescens as we understand it. We, therefore, feel justified in assuming that Mgwa is synonymous with Eudesme. JEgira, virescens (Carm.) S. and G. Plate 42, figs. 59, 60 Fronds very flaccid and gelatinous, 8-20 cm. (up to 45 cm.) high; branches arising from the main frond at wide angles, numerous, variable in length, usually enlarging above, blunt ; the ultimate ramuli short, arising almost at right angles; medullary filaments few to many very loosely bound together, composed of cells very variable in size even in the same filament, up to 55/* diam., 1.5-2.5 times as long as the diameter, constricted at the cross-walls, with numerous, small, disk- shaped chromatophores, giving rise to longer or shorter lateral, color- less, branched filaments bearing in turn the loose fascicles of free, branched, assimilating filaments, zoosporangia, and long colorless hairs, the latter terminating the colorless branches; cortical assimilating filaments slender, branched, tapering slightly at the apices, terminal ramuli 10-15 cells long, fascicled, arcuate, cells cylindrical below, becoming slightly moniliform above, 15-20/* diam. ; hairs numerous, long, terminating the lateral branchlets from the medullary filaments ; zoosporangia broadly ellipsoidal, short cylindrical, obovate to rhombic- ovate, arising near the outer ends of the colorless lateral filaments, 70-90/* (up to 120/*) long, 25-65/* broad; gametangia short, lateral and seriate, secund on the cortical filaments. Growing on rocks and on eel grass in the littoral and sublittoral belts. Shumagin Islands to Sitka, Alaska. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 11. Eudesme virescens Saunders, Alg. Harriman Exp., 1901, p. 423 (?); Setchell and Gardner, Alg. N.W. Amer., 1903, p. 249. Liebmannia sp. Saunders, loc. cit., p. 424, pi. 49 ( ?). Mesogloia virescens Carmichael, in Hooker, Engl. PL, vol. 2, 1833, p. 387. While we suspect that the plant referred by Saunders to Eudesme virescens and so referred to by us in our Algae of Northwestern America ( loc. cit.), may be the same species as that collected by one of us at Sitka, we have not been able to examine a specimen to make certain. Our plants show no gametangia but have zoosporangia and 548 University of California Publications in Botany [Vol.8 the vegetative structure of Eudesme J. Ag. A careful comparison of Wyatt's Mesogloia virescens, no. 49 (Alg. Danm.) with our various specimens shows so much variation in dimension of zoosporangia and peripheral filaments that we feel nearly safe in referring all to the European species which we refer, for reasons given above, to JEgira. 29. Meneghiniella S. and G. Fronds cylindrical, more or less profusely branched, flaccid and lubricous, with growth trichothallic ; medulla composed of numerous colorless filaments compactly coalescing, giving rise to an abundance of short, usually arcuate, color bearing, cortical filaments, whose lower branches are transformed into linear, plurilocular gametangia, with nearly uniseriate loculi, and usually more or less fasciculate ; zoo- sporangia uncertain. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 5. Cladosiphon J. Agardh, Till Alg. Syst., II, 1882, p. 40 (not of Kuetzing). The genus described above is founded on a plant growing in the Gulf of California. It seems, however, that our type must be cogeneric with the Liebmannia Posidoniae Meneghini, whose type locality is Naples and which was selected by J. G. Agardh as typical of Kuetzing 's genus Cladosiphon. We have shown above the results of our attempt to determine the type of JEgira, which have led us to the conclusion that the genus, as founded by Fries, includes the later genera published under the names of Cladosiphon, Castagnea, and Eudesme. Our plants seem to be typically trichothallic and have longer or shorter, slender, linear gametangia, which are, at times, very slightly swollen at, or about, the middle. The gametangia arise near the base of the cortical filaments and are more or less fasciculate. As we interpret Meneghini 's figures of Liebmannia Posidoniae (1843, pi. 5, fig. 1), he has represented the gametangia on the upper side of figure b and also in figure c. A confusion is introduced, however, in his representation of the contracted contents of the swollen terminal cells of the cortical filaments (figs, b, d, c, and /) by which they are made to seem to agree with the gametangia of Liebmannia Levillei J. Ag. The figures of J. G. Agardh (1882, pi. 2, figs. 3a, 3&) represent the cortical filaments with their swollen terminal cells and their game- tangia of his Cladosiphon zostericoJa of Australia which seems clearly to be referred to Meneghiniella. The genus Meneghiniella, in our esti- mation, included M. Brandegeei S. and G. (type), M. Posidoniae 192r, \ Setchell-Gardner : MeJanophyceae 549 (Menegh.) S. and G. (Liebmannia Poddonkie Meneghini, 1843, p. 300, pi. 5, fig. 1), M. zostericola (J. Ag.) S. and G. (Cladosiphon zostericola J. G. Agardh, 1882, p. 43, pi. 2, figs. 3a, 3&), and possibly M. erythram (J. Ag.) S. and G. {Cladosiphon erythraeum J. G. Agardh, 1848, p. 55). Meneghiniella Brandegeei S. and G. Plate 47, fig. 11, and plate 49, fig. 16 Fronds very slender, soft and flabby, 8-16 cm. high, up to 600/x diam., irregularly and alternately branched, with branches coming off at wide angles ; medullary filaments composed of large thin-walled, colorless cells up to 100/x diam. and several times longer than the diameter, firmly agglutinated, becoming much smaller toward the periphery; cortical assimilating filaments simple, clavate, free, not compact, more or less arcuate, composed of 7-12 moniliform cells; gametangia fasciculate with uniseriate loculi for the most part, 40-50^ long, 7-10/x diam. ; zoosporangia and hairs unknown. Habitat unknown. La Paz, Lower California. Setchell and Gardner, Phyc. Cont.. VII, 1924, p. 5. The specimens on which this species, as well as the genus, are founded are not as representative as could be desired, but the char- acters are plain both as to the structure of the frond and the shape, position, and the structure of the gametangia which can clearly be determined. The tips of the filaments indicate trichothallic growth. family 12. HETEROCHORDARIACEAE fam. nov. Main fronds erect, with main axis and lateral branches of limited growth ; growth in length subapical as in the Chordariaceae ; axes of three tissues: (1) medullary or central of elongated parallel, colorless cells; (2) intermediate of shorter colorless cells, and (3) cortical of short, anticlinal rows of colored cells, the distal cell being more or less swollen ; zoosporangia and gametangia on full-sized, similar plants. It seems something like violent procedure to separate this family from the Chordariaceae, but the presence of full-sized gametangial plants in this family and the lack of any such plants among the Chor- dariaceae (as limited by us) is certainly significant, in view of the more recent indications as to the relative characteristics of the gametophyte and sporophyte among the Melanophyceae. 550 University of California Publications in Botany [Vol. 8 30. Heterochordaria S. and G. Main fronds cylindrical, at first solid, later becoming hollow and at times slightly swollen, not forked but densely clothed on all sides with longer or shorter, subcylindrical or slightly flattened, ramuli. few to many arising from a thin prostrate, profusely branched or lobed. parenchymatous thallus firmly adhering to rocks; interior of fronds composed of thick-walled, colorless, parenchymatous cells surrounded by small color bearing surface cells in anticlinal rows; zoosporangia situated among numerous paraphyses and scattered hairs ; gametangia arising through the transformation of approximately the lower two- thirds of the cortical filaments. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 6. The finding of both zoosporangia and gametangia on macroscopic plants of Chordaria abietina Rupr., and the finding only of zoo- sporangia on the well known Chordaria jiageUiformis (Muell.) Ag. indicates the extremely strong probability of a fundamental difference between C. abietina and C. fiagelliformis as to the character of the gametophyte. In view of our present attitude as to the bearing of such a difference on the taxonomic position of the two plants, it seems desirable to make C. abietina the type of a new genus and a new family of the Ectocarpales and to place C. flagelliformis and its allies of the Chordariaceae (in restricted sense) in another order, the Chordariales and near to the Dictyosiphonales to which they (the Chordariaceae as we restrict them) seem fairly closely related. There may be associated with this family, another genus, or even two, viz., Ruprechtiella Yendo (Trav. Mus. Bot. de l'Acad. Imp. Sci. St. Petersb., vol. 10, 1913, p. 117) which may be the same as Analipus Kjellm., and the Chordaria Gunjii Yendo (Nyt. Mag. Naturvidensk., vol. 51, 1913, p. 280) which may simply be the gametangial form of some species of Myelophycus. Heterochordaria abietina (Rupr.) S. and G. Plate 36, figs. 18, 19, and plate 91 Erect fronds, 10-30 cm. high, flaccid, gregarious, many arising from a single, widespreading, profusely branched, parenchymatous horizontal thallus ; main axis slightly compressed above, terete below, thickly beset on all sides with radiating ramuli 1-5 cm. long, tapering at both ends, cylindrical to narrowly linear, both main axis and ramuli I 925 ] Setchell-Ganlner: MeUmophyceoi 551 solid when young, becoming more or less fistulose when old ; color dark olive brown ; erect fronds annual, base perennial ; zoosporangia ellip- soidal to obovoid, numerous, sessile on the parenchymatous cells or at times lateral near the base of the paraphyses, 50-60^ long, 30-40/* broad, paraphyses numerous, clavate, 70-85//. long, terminal cell much enlarged, pyriform to spherical; gametangia very compact, covering the whole surface of the ramuli, without paraphyses or hairs, pluri- seriate, terminating in a 2-3-celled, sterile, usually clavate tip ; plants dioecious. Common on rocks in the middle and lower littoral belts. Ranging from St. Lawrence Island in the Bering Sea to central California (near Point Conception). Setchell and Gardner, Phyc. Cont., VII, 1924, p. 6. Chordoma abietina Ruprecht, MS. in Farlow, List Mar. Alg. U. S., 1876, p. 357; J. Agardh, Till Alg. Syst., part 2, 1882, p. 74, pi. 3, figs. 2a, 2b; Saunders, Alg. Harriman Exp., 1901, p. 424; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 251, pi. 18, figs. 16, 17 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 281; Tilden, Amer. Alg. (Exsicc), no. 348; Farlow, Anderson and Eaton, Alg. Exsicc. Amer.- Bor., no. 94. Heterochordaria abietina (Rupr.) S. and G. is one of the very fre- quently occurring algae of western North America from Point Con- ception in California northwards. It has two striking peculiarities, viz., the possession of a well developed horizontal thallus from which the erect fronds arise and also of the gametophyte of the same size as the sporophyte. The horizontal thallus is often a fair sized and reasonably conspicuous structure, at times several centimeters in diameter, spreading at the periphery and consisting of overlapping narrow, blunt and emarginate lobes. From certain appearances, it seems possible that it may be perennial. In the sporophyte, the zoo- sporangia seem to be confined to the ramuli, and this seems to be the case also with the gametangia of the gametophyte which have not been seen to occur on the main axis. J. G. Agardh (1882, p. 74, pi. 3, figs. 2a, 2b) has figured and described the gametangia and Okamura (1910, p. 122, pi. 85, figs. 14. 15) has figured and described both the zoo- sporangia and the gametangia. We have previously (1903, pp. 251, 252, pi. 18, figs. 16, 17) figured and described the horizontal thallus. f)")^ University of California Publications in Botany [Vol.8 family 13. CHNOOSPORACEAE fam. nov. Fronds elongated, solid, slender, more or less compressed, more or less regularly dichotomous or proliferous from the margins, growth subapical, not trichothallic, of two sets of tissues, an inner of large cells, an outer of one or more layers of small colored cells in short anticlinal rows ; hairs in clusters arising from the bottoms of shallow pits ("cryptostomata") ; zoosporangia unknown; gametangia uni- seriate, arising about the clusters of hairs and spreading over the surface, at times into confluent sori. The genus Chnoospora. is usually ranked as of uncertain position ("incertae sedis"), but by some placed in the Encoeliaceae, a family now abandoned and with its members distributed and separated. In its sori, surrounding the hair clusters, it resembles closely those of CoJpomcnia, but in habit and terminal growth it is very different from that genus and its near relatives. The frond is not directly to be com- pared with that of any brown alga at present known to us. The slender, compressed and regularly dichotomous frond at once sets it apart. The subapical method of terminal growth relates it more nearly to Hetero- chordaria, but Chnoospora is known only with gametangia as are the members of the Scytosiphonaeeae. The method of subterminal growth is like that of the genus Stilophora, but it is very different from that genus in not having free cortical filaments. Chnoospora may be con- sidered to be a further development of the cylindrical Scytosiphon- aceae, with the loss of the terminal hair and the growth in length localized. It certainly seems best, at present, to place Chnoospora in a family of its own. The structure and the fructification were first described by Ethel Sara Barton (1897). Borgesen (1924, p. 264) calls attention to the resemblance of the growth point to those of Scytothamnus anostralis and Coilodesme buUigera. Both of these plants probably are to be referred to the Chordariales since they possess only zoosporangia of a distinct unilocular type and it is to be suspected that they have microscopic gametophytes. 31. Chnoospora J. Ag. Fronds cylindrico-compressed, solid, ecostate, profusely branched, attached by a solid holdfast ; branching dichotomous, fastigiate ; fronds differentiated into two tissues, the center of larger, colorless prismatic cells, elongated longitudinally, surrounded by two layers of small cells 1925] Setchell-Gardner : Melanophyceae 553 containing: the endochrome ; reproduction by plnrilocular gametangia surrounding: depressions containing: tufts of extruding hairs, and, by extension at times, nearly covering the fronds. J. G. Agardh, Nya Alg., 1847, p. 7. The type species of the genus is C. pacified from St. Augustin, [Mexico, listed later by J. G. Agardh as C. fastigiata var. pacified. Chnoospord is the only genus we feel inclined to refer to the family Chnoosporaceae, for the reasons given above in discussing the family. Chnoospora pacifica J. Ag. Fronds densely caespitose, many arising from a solid disk-shaped holdfast, of the same diameter throughout except at the slightly atten- uated apices, 2.5-7.5 cm. high, 2 mm. wide. 1 mm. thick ; branches forming acute angles. J. G. Agardh, Nya Alg., 1847, p. 7 ; Setchell and Gardner, Mar. Alg. Gulf Calif., 1924, p. 728. Chnoospora fastigiata var. pacifica J. G. Agardh, Sp. Alg., 1848, pp. 171, 172. Barton, On the fruit of Chnoospora fastigiata, 1898a:, p. 507. Sargassum piluliferum Collins, Holclen and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 537&. The material upon which this species, as well as the genus, was founded was collected by Professor F. M. Liebmann at St. Augustin, in the southern portion of the state of Oaxaca, on the Pacific coast of Mexico. This is a tropical locality and the Chnoospora was accom- panied by several other marine algae described also in the same paper. J. G. Agardh also described a species of Chnoospora, at the same time, from La Guayra on the Caribbean Sea which he named C. atlantica. In 1848, he united these two species as varieties under a new name, C. fastigiata. The name C. pacifica has the rights of priority and Ave have retained it. It seems widespread in the Pacific and Indian oceans, although possibly appearing in a number of forms. It seems likely that C. pannosa J. G. Agardh, from Hawaii, is simply a pannose growth form of C. pacifica. C. atlantica J. Ag. from the Caribbean has the fronds dilated under the axils and may be a distinct species. C. implexa Hering is doubtful and may very possibly prove to be of quite a different genus, but there are forms of our own coast species approaching it in habit. 554 University of California Publications in Botany [Vol.8 Order 3. DEHMARESTIALES ord. now Frond erect, filamentous or membranaceous, cylindrical to slightly compressed, ligulate or even broadly membranaceous; growth terminal, trichothallic ; frond clothed with densely crowded exserted, simple or branched, colored, monosiphonous filaments (hairs) when young or in the growing portion, these filaments early or tardily deciduous as age advances; tissues several; reproduction by zoosporangia ; game- tangia unknown, probably borne on a microscopic gametophyte. The possession of longer or shorter, exserted. colored, monosiphon- ous filaments, especially on young plants or younger portions of peren- nial plants, has always been considered as characteristic of Desmarestiu and the related genus Arthrocladia, It also is characteristic of the later described and very nearly related genus Phaeurus Skottsb. There have been several other algae with this permanent or partially transient covering of colored monosiphonous filaments, viz. ; Mesogloia Andersonii Farlow, Myriocladia Sciurus Harv., M. Chorda. J. Ag.. M . grandis Howe, M. callitricha Rosenv., M. Kuromo Yendo, and pos- sibly M . capensis J. Ag. and Mesogloia natalensis Kuetz. The first two seem to have been investigated by Kuckuck, who did not live to publish his results, but some of his notes and sketches were used by Oltmanns (1922, pp. 19-22). From these we learn that Kuckuck had assigned Mesogloia Andersonii Farlow (at least in part) and Myriocladia Sciurus Harv. to a new genus "Myriogloea" and had proposed for these a distinct family, presumably named Myriogloiaceae. Reserving the discussion of the different species until later, we may state that it seems best to adopt Myriogloiaceae, as of Kuckuck and assign it to our newly established order Desmarestiales, since the terminal growth is most typically trichothallic exserted hairs are prominent, mono- siphonous and colored, and the reproduction is typically only by zoosporangia. Key to the Families 1. Cortical filaments free 14. Myriogloiaceae (p. 555) 1. Cortical filaments coalescent into a tissue 15. Desmarestiaceae (p. 558) 1925] Setchdl-Gardner: Melanophyceax 555 family 14. MYRTOGLOIACEAE KUCKUCK Fronds filamentous, erect, cylindrical, solid or later tubulose, central strands of large colorless cells, peripheral of two sorts, free, short, cortical filaments with rounded terminal cell and exserted mono- siphonous filaments, with basal meristematic cells and outer cells pro- vided with phaeoplasts and. consequently colored; colorless hairs absent; terminal growth conspicuously trichothallic ; zoosporangia arising at the base of the cortical filaments, unilocular; gametangia unknown. Kuckuck, in Oltmanns, Morph. und Biol, der Algen, ed. 2, vol. 2, 1922, p. 19 (by implication). Myriogloeeae Oltmanns, loc. cit. The family thus far is credited with a single genus, Myrwglom, although if is far from certain that Mesogloia Andersonii Farlow is cogeneric with Myriodadia Sciurus Harv., as a glance at Kuckuck 's figures (loc. cit., p. 22, fig. 312, 1 and 2) will indicate. The very robust plants assigned to Myriodadia Sciurus Harv., 31. Chorda J. Ag., and M. grandis Howe may, or may not, belong to the same genus as Mesogloia Andersonii Farlow and Myriodadia callitricha Rosenv. We feel reasonably certain in referring the two last species to Myrio- gloia, which is absolutely distinct from Myriodadia J. Ag., as typified by 31. Lovenii J. Ag., designated type of the genus, at least if one credits Kuckuck 's sketches (cf. Oltmanns, 1922, p. 37, fig. 328 and p. 38, fig. 329) as having been drawn from correctly determined specimens. Influenced by these sketches and the various descriptions (notably that of Kylin, 1907, pp. 88-90, fig. 21), we are inclined to refer Myriodadia J. Ag., as limited to M. Lovenii, to the Myrio- gloiaceae as here established, as a second genus, distinguished by hav- ing its axis made up of a single longitudinal and persistent cell row (monosiphonous axis), whereas in Myriogloia this condition is soon lost, the frond becoming polysiphonous or hollow below, and in the thicker species {Myriodadia Sciurus, M. Chorda, and M. grandis) the medullary portion is probably polysiphonous from the first (cf. Kuckuck 's figure 312, 2, on p. 22 of Oltmanns, 1922) and remains solid. So far as our opportunities are concerned, there is insufficient material to follow out a detailed investigation and we cannot proceed farther than the above suggestions. Oltmanns, however, keeps Myrio- dadia (1922, p. 34) distinct and establishes a special tribe, Myrio- cladieae, for it under the family Spermatoehnaceae. Sperniatochnus, however, has a distinct apical cell. 556 University of California Publications in Botany [Vol.8 32. Myriogloia Kuck. Fronds cylindrical, more or less profusely branched, flaccid and lubricous, with distinct trichothallic growth ; medulla composed of numerous colorless, equally large filaments slightly coalescent, giving rise to numerous, narrow, long descending filaments and to an abun- dance of short, straight or uncinate, cortical, colored filaments and long, exserted, hair-like, assimilating filaments with cells densely con- gested with chromatophores; reproduction in macroscopic plaifts by zoosporangia borne among the cortical filaments. Kuckuck, MS (fide, Oltmanns, Morph. und Biol, der Algen, 1922, p. 19 et seq.). We have already explained the situation in regard to this genus and its position. It is closely related to Myriocladia but lacks the distinct and persistent monosiphonous axial filament. Its relation to the southern hemisphere Pacific species of Myriocladia is to be con- sidered, with the strong possibility that, upon careful study, they may be segregated generically. Myriogloia, in restricted sense, prob- ably includes three species : M. Andersonii (Farlow) Kuckuck {Meso- gloia Andersonii Farlow, Report U. S. Fish Comm. 1876, p. 715), M. caUitriclia (Rosenvinge) S. and G. {Myriocladia caUitricha Rosen- vinge, Gronl. Havalger, 1893, p. 855), and M. capensis (J. Ag.) S. and G. {Myriocladia capensis J. G. Agardh, Spec. Alg. vol. I, 1848, p. 54). If the Mesogloia natal ends Kuetzing is to be united with the last, the garnet angia ( ?) depicted in Kuetzing 's figure (Tab. Phyc, vol. 8, pi. 10, fig. II, e), if they really occur in the plant described and figured, call into question our disposition of this entire family. Myriogloia Andersonii (Farlow) Kuck. Plate 47, figs. 8-10, plate 48, figs. 12-14, plate 49, fig. 17, and plate 76 Frond very gelatinous, solitary or gregarious, cylindrical, 8-30 cm. (up to 50 cm. high), 1-3 mm. diam., profusely branched; branches mostly alternate, long and filiform, very slightly attenuated at the extremities, when young beset on all sides with relatively long, mono- siphonous, conspicuous filaments with band-shaped chromatophores, disappearing when older ; color dark olive brown ; cortical filaments moniliform, mostly uncinate, 6-9 celled, upper cells enlarged ; uni- locular zoosporangia developing at the base of the cortical filaments, uniformly distributed over the frond, pyriform, 36-45/* long, 18-21/* broad. 1925] Setcheli-Gardner: Mehnwphgceae 557 Growing on rocks in the lower littoral and upper sublittoral belts, or in tide pools higher up. The present known range is from Sitka, Alaska, to San Diego, California, Kuckuck, loc. cit. Mesogloia, Andersonii Farlow, New Alg. U. S., 1889, p. 9, pi. 87, fig. 2 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 250 ; Farlow, Anderson and Eaton, Alg. Exsicc. Amer.-Bor., no. 163; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 925. Chordaria flagelliformis Tilden, Amer. Alg. (Exsicc), no. 349. Much variation in the size of plants of Mijriogloia Andersonii has been noted in collections from different localities. Tilden 's plant (loc. cit.) is only about seven centimeters long and quite slender. No. 163 of Farlow, Anderson and Eaton (loc. cit.) is a young plant with abundance of hair filaments and is about twenty-five centimeters long. No. 925 of Collins, Holden and Setchell (loc. cit.) is an older plant, more robust throughout, about twenty centimeters long. Plants from Point Arena, California, in the Herbarium of the University of Cali- fornia, measure about fifty centimeters long. Likewise variations in the size of the mature zoosporangia are very noticeable, some collec- tions having them twice as long as those in others. The specimens distributed by Farlow, no. 163, 1881, are entirely clothed with brown exserted filaments. Farlow, however, in his description (1889, p. 9) states that it was supposed that all the specimens distributed were fertile and speaking of projecting filaments says: "In my specimens preserved in alcohol, they are not to be seen." We have been able to examine the alcoholic specimens referred to and find some, although scanty, exserted filaments. We find specimens of all condi- tions from being densely clothed with hairs to those with very few, but none in which they are entirely absent. We have tried to find out whether these various forms may be separated by any other characters than frequence or scarcity of exserted filaments, but can find none applying satisfactorily. We have, therefore, placed all of our Pacific coast material under the one species. Judging from what Oltmanns states from the imperfect notes in his possession, Kuckuck has two entities from the Mesogloia Andersonii Farlow, viz., Mijriogloia Ander- sonii, as stated above, and Chordaria Andersonii (cf. Oltmanns, 1922, pp. 20 and 23, fig. 313, 4). Of the latter, we have no further knowl- edge, although Oltmanns says "Chordaria Andersonii Kuckuck mscr. (Mesogloia Andersonii Farlow) . " There is nothing in Farlow 's mate- rial, either dried or in alcohol, which corresponds in methods of growth to Kuckuck 's figure. 558 University of California Publications in Botany [Vol.8 family 15. DESMARESTIACEAE kjellm. Fronds erect, sparsely branched in some species, profusely branched in others, from cylindrical through compressed to broadly membran- aceous, solid, arising' from a simple monosiphonous branched filament of trichothallic growth in length, soon becoming complex through the downward growth of closely applied branching corticating filaments which results in the production of several distinct tissues; central monosiphonous and branching filament remaining distinct, although, at times, becoming obscured through thylose growth filling the cavities of the large cells; tissues outside the central monosiphonous axis, various, usually large celled and more or less regular, the outer being smaller and forming a distinct outer tissue of one, or few, layers of cells; zoosporangia known in only a few species scattered or in small patches, arising by direct transformation of the outer cells ; plants mostly undergoing change of color (to verdigris green) and odor ("sharp") soon after being removed from the water and exercising a bleaching effect on other algae with which they come into contact. Kjellman, Enum. PI. Scand., 1880, p. 10, Alg. Arctic Sea, 1883, p. 261. The Desmarestiaceae form a very distinct group and one which is readily separable from that of the Myriogloiaceae. It represents a very distinct type of frond development, at least so far as the final product is concerned. There are two centers of distribution, viz., North Atlantic and North Pacific as contrasted with the Antarctic and Subantarctic regions, and it is doubtful whether there are any species common to the two centers. There are, at present, referred to the Desmarestiaceae three genera, Desmarestia, the type genus, Arthrocladia, by some referred to a separate family, and the Phaeurus of Skottsberg. The first is from both Arctic and Anarctic centers, the second is North Atlantic (European and American) while Phaeurus is from Fuegia. While the Myriogloiaceae are seemingly the more primitive and related to such forms as ^Egiraceae, the Desmarestiaceae are more complex, at least, with Phaeurus as the more simple and the ligulate Desmarestias as the more complex. 1925] Setrh el I -Gardner: Melanophijceae 559 33. Desmarestia Lamour. Fronds attached by a solid, parenchymatous disk, exceedingly variable in size, with a longer or shorter, usually cylindrical stipe, more or less profusely branched, wholly or in part terete or slightly flattened or decidedly complanate and ligulate to broadly expanded and membranaceous, with or without a midrib; branching distichous and either wholly opposite or alternate, or both opposite and alternate ; fronds composed of a single fundamental axial filament of cylindrical cells and more or less profusely branched, colored and projecting beyond the surface at the apices and along the margin and in the juvenile stage, early deciduous, with trichothallic growth, sur- rounded in the frond by one or more layers of large colorless cells enclosed by a cortical tissue of small assimilating cells; reproduction by zoosporangia composed of slightly transformed surface cells, in conspicuous. Lamouroux, Essai, 1813, p. 23. The genus Desmarestia may be divided into three groups, some- times reckoned as distinct genera, or at least it may be divided into two genera. The cylindrical or at most only slightly flattened species with mostly opposite branching, constitute Dichloria, while the flat- tened forms are retained under the name of Desmarestia, or the sub- genus Eudesmarestia, with branching mostly alternate. The Dichloria group has the hairs scattered as in Phaeurus, but unlike that genus, soon loses them. In the Eudesmarestia group, the hairs are branched and are the very tips of incipient branches, some, or even many, of which do not develop into permanent structures. In Phaeurus the hairs are scattered, but are more permanent and per- sistent for a longer time than in the Dichloria group. In Arthrocladia, the hairs are in whorls, branched, persistent, and possibly represent the tips of incipient branches. One well-known characteristic of the species of Desmarestia may well be emphasized and that is their habit of turning verdigris green on drying. This color is associated also with their power to bleach other algae in contact with them. The change of color is followed, sooner or later, by decomposition. This happens quickly in the more slender species of the Vir ides-section and fairly early in the members of the Herbaceae-section, but affects only the younger and more delicate portions of the members of the Aculeatae- section. 560 University of California Publications in Botany [ VoL - 8 The genus Desmarestia is well represented on the Pacific coast of North America, extending in its various species from the Bering Sea on the north to Santa Catalina Island off the coast of southern Cali- fornia on the south, so far as is known of its distribution at present. The species of both the ligulate and the filiform groups manifest a degree of variability which has up to date made it exceedingly difficult to determine specific limits. It is evident from the relatively good supply of herbarium material at our disposal, that more than mere superficial resemblances must be resorted to in order to make a satis- factory disposition of the group. Knowledge of the entire life-history is highly essential. Little is known of reproduction. The character of the hairs, present only in the early stages of growth — absent in most of our herbarium specimens — is, so far as our experience shows, of marked constancy and would doubtless be of decided assistance in classification. The histological structure, very difficult to restore after the material has once been thoroughly dried, certainly will prove of decided value in establishing specific entities when studied in fresh material. "We have made use of all of these different phases as far as the material and time at our disposal permits and have come to the follow- ing conclusions regarding the number, arrangement, and distribution of the Pacific coast species. Key to the Species 1. Fronds filiform and either terete or compressed 2 1. Fronds ligulate or membranaceous (Herbaceae) 6 2. Fronds terete or in part subterete, branching mostly opposite (Virides) .... 3 2. Fronds compressed or subterete below, branching mostly alternate (Aculeatae) 5 3. Fronds profusely branched; tylloses few or wanting 4 3. Fronds moderately branched; tylloses filling central tube to near tip 3. D. farcta (p. 562) 4. Branches long, attenuated, successive orders gradually reduced in s i ze 1. D. media (p. 561) 4. Branches slightly attenuated, blunt, successive orders abruptly changing to smaller size 2. D. pacifica (p. 562) 5. Fronds up to 15 dm. high and 3 mm. wide, with aculeate margins, moderately branched 4. D. latifrons (p. 563) 5. Fronds up to 7 dm. high and 1.5 mm. wide, with smooth margins and profusely branched 5. D. intermedia (p. 564) 6. Fronds unbranched or with a few short branches at the base 10. D. foliacea (p. 569) 6. Fronds regularly branched 7 7. Fronds membranaceous, up to 1 m. wide 9. D. latissima (p. 568) 7. Fronds ligulate, up to 1 dm. wide 8 8. Fronds more than 3 m long with prominent aculeae along the margins 8. D. munda (p. 567) 8. Fronds less than 3 m. long, smooth or with fine aculeae along the margins.. 9 1925] Setchell-Gwrdner: Melanophyceae 561 9. Fronds up to 8 dm. high, 6-8 mm. wide, without distinct lateral veins, stipe cylindrical, 3-4 mm. diam 6. D. ligulata (p. 566) 9. Fronds up to 2.5 m. high and 2 cm. wide, with distinct lateral veins, stipe flattened and short 7. D. herbacea (p. 566) Section 1. Virides Desmarestia, subgen. Dichloria (Grev.) DeToni. Fronds cylindrical or only very slightly compressed ; branching mainly opposite. DeToni. Syll. Alg., vol. 3, 1895, p. 456. Dichloria Greville, Alg. Brit., 1830, p. 39. 1. Desmarestia media (Ag.) Grev. Fronds cylindrical, considerably cartilaginous, branching very profusely and opposite throughout ; color dark brown, nearly black on drying; branches of 5-6 orders, each successive order gradually reduced in size, composed of an axial row of relatively large cells sur- rounded by 2-3 layers of large, more or less irregular, elongated cells surrounded in turn by a single palisade layer of cortical assimilating cells, the terminal ramuli long-attenuate and acute. Growing on rocks in the upper sublittoral belt. Unalaska, Alaska. Greville, Alg. Brit,, 1830, p. xl; Pease, P. S. M. S. Publ., vol. 1, 1917, p. 386. Desmarestia aculeata f. media' Setchell and Gardner, Alg. N.W. Amer., 1903, p. 246; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1036. Sporochnus medius Agardh, Sp. Alg., vol. 1, 1820, p. 153, Icones Alg. Ined., 1821, pi. 16. Desmarestia viruiis Saunders, Alg. Harriman Exped. 1901, p. 422 ( ?). Desmarestia media var. tenuis S. and G. Fronds cylindrical and with opposite branching throughout, very profusely branched, 3-4 dm. high, stipe 2-3 mm. diam., ultimate ramuli very slender and flaccid, 200-250/x diam. ; axial filament of the fronds very large, 100-1 25/^ diam., surrounded by a single layer of large, colorless cells between the axial filament and the small cortical cells. Growing on stones, etc., in the upper sublittoral belt. Alaska (Juneau) to Puget Sound, Washington. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 7. 562 University of California Publications in Botany t VoL - 8 There has been considerable confusion as to this species which was described by C. A. Agardh from specimens said to have been collected at Unalaska by Chamisso. Our specimens were collected at Unalaska and in general seem to answer definitely to the description and speci- mens of Agardh. The fronds are nearly exactly cylindrical through- out, the branching is almost entirely opposite and the branches decrease in size gradually for each successive order of ramifications. In habit, branching, and in structure, D. media is clearly related to D. viridis, but is a more ample and a coarser plant. It is clearly distinct in habit, branching and structure from any of the aculeata- group. Cross-sections do not show the dense external palisade-layer of cells of the D. acideata and its allies. The variety tenuis is typically a much more slender plant, but is, nevertheless, coarser and more rigid than any of the forms of D. viridis. The variety seems to be more southern in its range than the type of the species. 2. Desmarestia pacifica S. and G. Fronds slightly compressed below, terete above, branching opposite throughout, each succeeding order of branches much reduced in size ; subterminal branches 5-8 cm. long, tapering upwards but very slightly, clothed with many, mostly opposite, blunt, ultimate ramuli, 2-5 mm. long; axial filament 40-50//, diam., surrounded by two layers of larger cylindrical to subcylindrical cells; cortical layer not definitely delimited. Santa Catalina Island, southern California. Setchell and Gardner, Phyc. Cont., VII, 1924, p. 6. We have only a single and imperfect specimen upon which to base this new species, but, while it closely resembles Desmarestia media, each successive order, the branches of which do not taper appreciably upwards, gives the plant a characteristic seemingly very distinct. 3. Desmarestia farcta S. and G. Fronds cylindrical throughout, at least 5 dm. high ; branches long, opposite throughout, not attenuated, blunt, of 3-4 orders, 0.5-1 mm. diam.; stipe and holdfast unknown; primary cylinder composed of cylindrical cells 20-25/* diam. in the center of a distinct "stele" of 3-4 layers of larger cylindrical cells 35-45/t diam., these surrounded by 2-3 layers of large colorless cells irregular in form, 100-125/i diam., 1925] Setchell-Gardner: Melanophyceae 563 and surrounded in turn by the cortex of colored, radially elongated cells ; reproduction and hairs unknown. Cast ashore. Argyle, San Juan County, Washington. Setchell and Gardner, Phyc. Cont,, VII, 1924, p. 7. We have but two portions of what may be a single ample plant of this species upon which to base this description. The character of the holdfast and of the lower parts of the frond cannot be chronicled at this time, but the characters of the parts which we have are so pro- nounced as to leave little doubt that the species has hitherto been overlooked and is undescribed. One of the incomplete specimens, probably a portion of the same plant, is in the collection of the Marine Biological Station at Friday Harbor, Washington, and we are able to examine it through the courtesy of Professor T. C. Frye. The densely "stuffed" condition of the cells of the central monosiphonous axis begins within two or three cells of the tip of each axis through the growth of tylloses, a condition not noted by us in any other of the specimens available to us. Section 2. Aculeatae Desmarestia, subgen. Eudesmarestia DeToni. Fronds compressed to slightly flattened, never strictly flattened- foliaceous, foliaceous, or cylindrical ; branching alternate, occasionally opposite below. DeToni, Syll. Alg., vol. 3, 1895, p. 457 (in part). 4. Desmarestia latifrons (Rupr.) Kuetz. Plate 90 Fronds of a firm, rigid consistency, dark brown in color, moderately branched, flattened throughout except the base of the stipe, several main lateral branches often exceeding in length the central axis, midrib manifest only in the lower older parts, 8-15 dm. high, 2-3 mm. wide; branches of 3-4 orders, those of the fourth order foliaceous above, reduced to small aculeae below or entirely absent ; hairs oppo- site, branched, arising from every cell of the central axial filament, closely congested along the entire margin of the corticated frond and of the ramuli except at their stipitate bases. Growing on rocks in the lower littoral and the upper sublittoral belts. Central Oregon (Coos Bay) to central California (Point Sur). 564 University of California Publications in Botany [Vol. 8 Ruetzing, Tab. Phyc, vol. 9, 1859, p. 40, pi. 95; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc.), no. 279; Farlow, Anderson and Eaton, Alg. Exsicc. Amer-Bor., no. 121. Spinularia latifrons Kuprecht, Tange, 1851, p. 375. The type locality of D. latifrons is, according to Ruprecht (loc. cit.)., Fort Ross, California. We have not seen any specimens from the type locality, although we have both collected algae in that vicinity. The species, as at present known, does not seem to be very abundant, but is scattered sporadically from Coos Bay, Oregon, to Point Sur, Monterey County, California. As far as our experience goes, it is one of the most stable, least variable species of the genus, and can readily be told from all other known species. The hairs are relatively long, very abundant and remain on until the plants are nearly maximum growth. These are replaced in part by numerous prominent teeth or aculeae at maturity. The fronds are decidedly rigid and of a dark color. Desmarestia latifrons resembles Desmarestia intermedia more than any other of the Acideatae, but is definitely flattened rather than com- pressed, and is not at all either membranaceous or foliaceous. It seems to be a very distinct species of the central Californian coast (extending over into the southern Oregon coast at Coos Bay), but is scarce even in its own territorj^. The species is related to D. aculeata (L.) Lamour., but more closely to D. intermedia P. and R., but is broader and more flattened than either. 5. Desmarestia intermedia P. & R. Fronds subcoriaceous, up to 6 dm. high and 1.5 mm. wide, dark brown in color, almost black on drying, very profusely branched, sub- opposite and fasciculate below, strictly alternate above ; stipe and lower parts cylindrical to subcylindrical ; ramuli decidedly compressed and lanceolate, densely clothed in the juvenile stage with relatively long hairs with alternate or opposite branching ; frequently, especially in the lower parts and in the older plants, two or more branches arising at the same node and on the same side of the frond. Growing on rocks in the upper sublittoral belt. Bering Sea to Puget Sound. Posters and Ruprecht, Illus. Alg., 1840, p. 13, pi. 26 (Excl. Syn- onymy) ; Desmarestia aeideata Setchell and Gardner, Alg. N.W. Amer., 1903, p. 246 (in part); Tilden, Amer. Alg. (Exsicc), no. 352. Possibly, Saunders, Alg. Harriman Exp., 1901, p. 422. 1925] Setchell-Gardner: Melanophyceae 565 Desmarestia intermedia has usually passed for D. acideata, true representatives of which do not seem to occur on our coast. D. acideata is characterized, at least in typical forms, by having its shorter and even longer branches margined by short branches of limited growth, asuming the form of spines. D. intermedia, however seldom has what appear to be genuine aculeae, i.e., short rigid spine-like branches of very limited growth such as are characteristic of D. aculeae and D. latifrons, usually there being little, if any, appearance of this sort. One characteristic of D. intermedia is the fasciculate branching below. The groups of branches springing from one point are subopposite, i.e., arise near one another, but one bunch is inserted higher on the axis whence both arise and not exactly on the opposite side. The plants we have assigned to this species vary from fairly broad (type) to medium (var. fuscescens P. and R., lac. cit.) and on to the most slender and almost terete forms (var. teretifolia P. and R., lac. cit.). Postels and Ruprecht (lac. cit.) seem to feel certain that the nomen nudum of Mertens, Fucus pseudoaculeatus, refers to their plant, hav- ing probably examined a specimen, but they consider the other three quotations, viz., Sporochnus medius Ag., Trmitaria confervaides Bory, and Desmarestia media Grev., as dubious. The first and the last refer, in our opinion, to what we call Desmarestia media. The middle plant is presumably another species, having been assigned to a plant of the southern hemisphere. Section 3. Herbaceae Desmarestia, subgenus Eudesmarestia, De-Toni (loc. cit.), in part. Fronds ligulate to broadly membranaceous or foliaceous; branch- ing abundant to very sparse or none, opposite or subopposite. De-Toni, Syll. Alg., vol. 3, 1895, p. 457. There is a very considerable variation in length, width, and branch- ing in our west American plants of the Herbaceae and there are very considerable difficulties in the way of classifying them either as dis- tinct species or as varieties. Miss Vinnie Pease (1917 and 1920) has made an effort in this direction and we have considered her account and the opinions she has expressed therein, very carefully. As a result of our study of the literature and of the specimens, we have decided to mark off the more striking plants of our coast as more or less distinct species. 566 University of California Publications in Botany [Vol. 8 6. Desmarestia lignlata ( Light f.) Lamour. Plate 87 Fronds plumose, up to 8 dm. high (in our territory), 6-8 mm. wide, firmly attached by a relatively large disk; primary axis closely beset with opposite, primary branches throughout, exceedingly variable in size and complexity of branching, some of which extend beyond its apex ; ramuli all opposite, closely crowded, of 4-5 orders, the ultimate reduced to fine teeth or aculeae, all very decidedly flattened ; midrib barely discernible in the primary axis and main laterals ; hairs very abundant, with short, very acute, opposite branches, giving rise along the margins of the ultimate branches to numerous small acute ramuli, many of which disappear at maturity. Growing on rocks in the upper sublittoral and lower littoral belts among and in the vicinity of the San Juan group of islands, Washington. Lamouroux, Essai, 1813, p. 25 ; Harvey, Phyc. Brit., vol. 1, 1847, pi. 115 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 247 ; Pease, P.S.M.S. Publ., vol. 1, 1917, p. 388, and vol. 2, 1920, p. 314 ; Wyatt, Alg. Danmon., no. 55 ; Le Jolis, Alg. mar. Cherb., no. 268 ; Hauck and Richter, Phyk. Univ., no. 420. Fucus ligulatus Lightfoot, Flora Scotia, vol. 2, 1777, p. 946, pi. 29. There may be still a question as to whether typical Desmarestia ligulata occurs on our coast. Pease (1920, p. 316) has argued that it does and we have adopted her opinion. There does occur on the island of San Juan and apparently in abundance a narrow, branched plant with short filiform stipe without prominent midvein, and lacking lateral veins, which seems to correspond to Lightfoot 's figure of Fucus ligulatus and description and with British plants. The type locality is on the Scottish coast and is probably on the Firth of Forth. 7. Desmarestia herbacea (Turner) Lamour. Plate 88 Fronds ligulate, 1.5-2.5 m. long, 1-2 cm. wide, the rhachis and the primary branches with a distinct, conspicuous midrib ; stipe very short, almost immediately flattened ; branches of 3-4 orders, each order much reduced but exceedingly variable in size, the primary branches often reaching the size of the rhachis, but interspersed with branches only a few centimeters long ; all of the branches, though 1925] Setchell-Gardner: Melanophyceas 567 relatively wide, tapering at the base to a small, distinct, almost cylin- drical connecting stipitate portion, and rounding at the apices; the branches exceedingly abundant in the juvenile stage, and all densely clothed at the apices and along the margins Avith acute hairs, the main terminal hairs with profuse opposite branching. Growing in tide pools in the lower littoral and upper sublittoral belt. Kodiak Island, Alaska, to southern California (La Jolla). Lamouroux, Essai, 1813, p. 25; Kuetzing, Tab. Phyc, vol. 9, 1869, pi. 100; Pease, P.S.M.S. Publ., vol. 1, 1917, p. 389; Ibid., vol. 2, 1920, p. 340, pi. 54, fig. 3, pi. 60, figs. 1-5. Desmia herbacea Postels and Ruprecht, Illus. Alg., 1840, p. 13. Desmarestia ligulata var. herbacea Collins, Holden and Setchell, Phyc. P^or.-Amer. (Exsicc.), no. 79b. Desmarestia ligulata Farlow, Anderson and Eaton, Alg. Exsicc. Amer.- Bor., no. 166. Fucus herbaceus Turner, Hist. Fuc, 1809, pp. 77, 78, pi. 99. The exact locality from which the material forming the type of the species of D. herbacea was collected is not known. Turner (loc. cit.), says, "North-west coast of America. Mr. Menzies." There is much variation in the size of the fronds as they occur along our coast. Turner says, "two feet or more long." We have seen specimens two and a half meters long. Some specimens, apparently mature, from La Jolla and from Pacific Grove, California, in our herbarium are almost as narrow as the specimens from San Juan County, Washington, which we have referred to D. ligulata. Other specimens approach very nearly to the narrow forms of D. munda as we have limited that species. However, the typical and abundant specimens are quite distinct from either of these species. They are midway between the two species as regards the number and character of the branches. Typical plants also show distinct, often heavy, lateral veins, at least on all the broader axes. 8. Desmarestia munda S. and G. Plate 89 Fronds attached by a firm parenchymatous disk, ligulate, relatively rigid and coriaceous, up to 8 m. long, 4-10 cm. wide, glossy yellowish brown in color, sparingly branched at maturity ; midrib prominent in the stipe and lower parts, becoming very inconspicuous or appearing only as a mere nerve above ; branches usually of 2 but in part 3 orders, the largest primary branches arising near the base at times as long as the central rhachis, and even wider in part, tapering rather abruptly 568 University of California Publications in Botany [Vol.8 to ;i small cylindrical connection at the base, acuminate or rounded above, the margins of all bearing very prominent, rather distant spine- like projections with more or less rounded angles above; stipe flattened almost to the base. Growing principally on rocks, just below extreme low water level down to ten fathoms or more in depth. From Puget Sound, Wash- ington, and probably considerably farther north, to southern Cali- fornia (San Pedro). Setchell and Gardner, Phyc. Cont., VII, 1924, p. 7. Desmarestia ligulata var. herbaeea Setchell and Gardner, Alg. N.W. Amer., 1903, p. 247; Tilden, Amer. Alg. (Exsicc), no. 244; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. LXXIX A (not B). Des- marestia latissima Pease, P.S.B.S. Publ., vol. 2, 1920, p. 319 (in part). Desmarestia munda is a long and moderately broad plant, with indistinct midvein, usually faint above, and lateral veins branching and disappearing before reaching the margins. The texture of the living plant is firm and its surface is smooth, shining, and usually free from epiphytes and epizoa. It is for this reason that we have chosen the specific designation. Pease (1920, p. 319) has published a Desmarestm latissima as of Setchell and Gardner and states that the type is in the Herbarium of the University of Minnesota. We are not responsible for the type, which we have not yet seen, but from the description in Pease, conclude that it includes D. munda as we estab- lish it, as well as a plant to which we did assign the provisional name D. latissima. The latter is the plant to which we limit the name, D. latissima. In our opinion, our D. latissima is sufficiently distinct from D. munda to justify keeping them separate. 9. Desmarestia latissima S. & G. Fronds membranaceous, broadly expanded, 3-4 m. long, up to 12 dm. wide, sparsely branched, exceedingly fragile; branches similar to the main axis, of one order only, or possibly at times a few small secondary branches, abruptly tapering at the base to a very slender connection with the main frond, rounded at the outer end, midrib and lateral veins inconspicuous. Growing in quiet water in bays and on mud flats, in the upper sublittoral belt. San Juan County, Washington. Setchell and Gardner, in Pease, P.S.B.S. Publ., vol. 2, 1920, p. 319 (in part). 1925] SeteheU-Gardner: Mclanophyceae 569 This is by far the largest species of Desmarestia yet discovered, and seems, as far as known to us, to inhabit relatively quiet water, at times merely resting on mud flats where the water is rarely disturbed except as it rises and falls with the tidal movements. One of its dis- tinguishing characters is its very large size. The most noticeable feature about it, however, is its exceedingly fragile nature. It is very crisp and brittle, wholly unable even to bear more than a fraction of its own weight without being ruptured. It is further characterized by the paucity of branches, as a rule bearing only a few primary laterals. This plant fully merits the name Desmarestia latissima, as already given by us, and is a plant of entirely different color, texture, habit, and place of growth from our D. munda. 10. Desmarestia foliacea Pease Fronds membranaceous, unbranched, attached by a small disk, tapering below to a very delicate cylindrical stipe, up to 1 m. long, 15 cm. wide, margins irregularly sinuate with occasional sinuate teeth, midrib evident, with opposite branches which break up into fine veinlets. Growing in the upper sublittoral belt. San Juan County, Wash- ington. Pease, P.S.B.S. Publ., vol. 2, 1920, p. 322, pi. 58, figs. 5-10, pi. 61, figs. 1-5. Desmarestia tabacoides Pease, P.S.B.S. Publ., vol. 1, 1917, pi. 84, figs. 2-7. We are not acquainted with this plant and know it chiefly through the above mentioned publications. There is a plant in the Herbarium of the University of California (Gardner, no. 2321) dredged in ten fathoms off Canoe Island, San Juan County, Washington, which may possibly belong here. The plant was attached to a clam shell by a relatively small, flat disk, has a small cylindrical stipe, a blade about four meters long and three decimeters wide. There are, however, a few small, short branches arising near the base, differing in this respect from the type material as described by Miss Pease. Through the kindness of Miss Pease, we have received, since writing the foregoing, two sheets of specimens of Desmarestia foliacea from Professor Josephine Tilden, of the University of Minnesota. The species seems closely related to D. taba-coides Okam., but differs in size, shape, lobing and venation. 570 University of California Publications in Botany [Vol.8 Order 4. CHORDARIALES ord. nov. Fronds erect, cylindrical, branched or unbranched, arising from a diseoid holdfast or from an expanded and lobed horizontal thallus, -with terminal, but subapical, growth, of three sets of tissues, central core of vertical rows of elongated cells (often disappearing leaving the erect filaments hollow), an intermediate layer of cells slightly longer than broad or thick, and an outer layer of longer or shorter assimilating filaments, more or less distinct or combined into a cortical tissue, near, or at, whose bases the zoosporangia are formed ; zoo- sporangia immersed among the assimilating filaments ; colorless hairs generally present ; gametangia and gametophytes unknown, the later presumably microscopic. In pursuance of the idea stated elsewhere, that those groups of Melanophyceae closely related in general vegetative structure and par- ticularly with the same method of growth in length in which no member is known to occur with gametangia, are to be suspected of possessing microscopic gametophytes, we have established the order Chordariales with the family Chordariaceae as the typical family, which includes Chordaria, typified in Chordaria flageUiformis (Muell.) Ag., as the type genus. The method of growth is that designated as subapical, as distinguished from trichothallic (e.g.,Ectocarpales and Desmarestiales) on the one hand and truly apical (e.g., Sphaeelariales, Dictyo- siphonales, etc.) on the other. The meristematic cells are situated in short apical filaments similar to the external assimilatory filaments, but are not terminal in those filaments, being situated two or three, or occasionally more, cells below the terminal cell. In other respects the structure approaches that of the iEgiraceae and the Myriogloiaceae where, however, the growth is trichothallic and both macroscopic sporophytes and macroscopic gametophytes are known. From Hetero- chordariaceae, the families of this order differ in lacking, so far as known, a macroscopic gametophyte. While the family Chordariaceae can be assigned to the order Chordariales, as the type family, yet the close relationship of the genera Coilodesme and Scytothamnus lead us to add a family for the first (Coilodesmaceae) and one for the second (Scytothamnaceae), although we have no members belonging to the latter within our territory. Key to the Families 1. Assimilating filaments distinct 16. Chordariaceae (p. 570) 1. Assimilating filaments united into a cortical tissue ...17. Coilodesmaceae (p. 577) 1925] SetcheH-Gardner: Melanophyceae 571 family 16. CHORDARIACEAE reichenb. (in part) Fronds erect, cylindrical, branched or simple, arising from discoid holdfasts or from lobed and expanded horizontal thalli, composed of three tissues or layers, an inner, of elongated, usually slender fila- ments (often disappearing and leaving a hollow center), an inter- mediate, of large, rounded, or somewhat elongated cells, and an outer cortical layer, of short, anticlinal, assimilating filaments, distinct from one another, but enclosed in an enveloping jelly ; zoosporangia immersed among the cortical filaments ; gametangia and micro- scopic (?) gametophyte unknown. Reichenbach, Conspectus Regn. Veg., 1828, p. 25 (in part; fide Pfeiffer, Nomen. Botan., vol. 1, 1873, p. 732, et auct. var.) ; Harvey, Ner. Bor. Amer., I, 1852, p. 121 (in part). Chordarieae Agardh, Syst. Alg., 1824, p. xxxvi (in part). The family of the Chordariaceae, while actually very distinct from all others of the Melanophyceae except that of the Heterochordariaceae because of its subterminal growth and frond structure, has been the recipient of several families in which the growth is clearly tricho- thallic. We may assume with reason, we think, that -the subapical method is intermediate between the apical and the trichothallic yet distinct from each of them. The outer, cortical, layer of short, but distinct, anticlinal filaments distinguish this family from both the Coilodesmaceae and the Scytothamnaceae. Key to the Genera 1. Fronds simple (in our species) or very sparingly branched 2 1. Fronds, typically, much branched 34. Chordaria (p. 571) 2. Fronds composed of a flattened, sterile part and an erect, fertile part 35. Analipus (p. 575) 2. Fronds not composed of two parts 36. Gobia (p. 576) 34. Chordaria Ag. (emend Grev.) Fronds cylindrical, filiform, branched, more or less rigid and cartilaginous to slightly lubricous, solid or at times becoming fistulous in age, composed of two tissues, a medulla and a cortex ; growth sub- apical ; medulla composed of relatively large, colorless, longitudinally arranged, firmly agglutinated filaments, with numerous small fila- ments, especially abundant in the center of the frond, interspersed ; cortical assimilating filaments short, firmly agglutinated ; reproduc- tion in macroscopic plants by zoosporangia borne among the cortical filaments. 572 University of California Publications in Botany [Vol. 8 Agardh, Syn. Alg. Scand., 1817, p. xii and p. 12, Jim. mut.; Greville, Alg. Brit., 1830, p. 44. The name Chord aria was first proposed by Link in 1809 to include a series of different plants, none of which, however, belong to this entity as now constituted. C. Agardh (1817, p. xii) modified the limits of the genus to include C. rotunda, C. divaricata, C. flagelli- formis, C. cabrera, and C. Filum and three other doubtful species. Of these C. flageUiformis is the only northern species of the series now generally included under the genus, but some authors also include C. divaricata, Greville (1830, p. 44, pi. 7) was the first definitely to establish the present limits of the genus by pointing out a type, viz., C. flageUiformis. Key to the Species 1. Fronds flagelliform, cylindrical throughout 2 1 Fronds irregular in diameter, flattened at the branching .3. C. dissessa (p. 574) 2. Fronds relatively robust, up to 9 dm. high, usually with many long, slender, primary branches 1. C. flageUiformis (p. 572) 2. Fronds slender, up to 3.5 dm. high, primary branches patent, scattered along the whole main axis up to 4 cm. long 2. C. gracilis (p. 573) 1. Chordaria flageUiformis (Muell.) Ag. Fronds attached by a small disk, solitary or in groups, lubricous, solid, 2-5 dm. (up to 9 dm.) long, branching close to the base into many branches of nearly equal length, or the main axis percurrent giving rise to numerous alternate branches, shorter above; branches usually of 3 orders; peripheral assimilating filaments 6-8 cells long, with terminal cell nearly spherical, and with numerous hyaline hairs interspersed ; color dark olive brown, nearly black ; zoosporangia ovoid to pyriform, arising at the base of the cortical filaments. Growing on rocks in the upper sublittoral belt. Agardh, Syn. Alg. Scand., 1817, p. 12 and p. xii. Fucus flageUi- formis Mueller, Flor. Dan., 1771, pi. 650. This species in its various forms seems to be exclusively boreal, though extending much farther south on the Atlantic shore than on the Pacific shore of North America. It is very common on the American and European shores of the North Atlantic. It has been reported as far north as the Spitsbergen Islands and in the Baltic Sea. The following forms have been reported within our territory. I92r>] Setchell-Gardner: Melanophyceae 573 Chordaria flagelliformis f . typica Kjellm. Plate 39, fig. 44 Fronds much branched, primary branch arising near the base, usually longer than the others, main axis nearly percurrent. Growing on rocks in the upper sublittoral belt. Bering Sea (St. Lawrence Island) to Sitka, Alaska. Kjellman, Om Spets. Thall. II, 1877a, p. 28; Reinke, Atlas, 1889, pi. 39, figs. 1-7 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 250 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 1037. Chordaria flagelliformis f. ramusculifera Kjellm. Fronds cylindrical, up to 30 cm. long, slender, with scattered, short, simple branches 1-1.5 cm. long, arising nearly perpendicular to the main axis ; cortical filaments 5-6 cells long, distinctly clavate, terminal cell moderately swollen ; zoosporangia elliptical, 55-60/x long, 25-30^ wide. Growing in the sublittoral belt. St. Lawrence Island, Alaska. Kjellman, Om Spets. Thall. II, 1877a, pp. 29, 30, pi. 1. figs. 10-12 : Setchell and Gardner, Alg. N.W. Amer., 1903, p. 251. Chordaria flagelliformis f. chordaeformis Kjellm. Fronds simple or sparsely branched ; cortical filaments nearly cylindrical with terminal cell slightly swollen ; zoosporangia 60-70^ long, 25-30/x broad. Kjellman, Om Spets. Thall. II, 1877a, pp. 28, 29, pi. 1, figs. 13-15 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 251. This form has been reported by Kjellman in the upper sublittoral belt at Konyam Bay, Siberia, and may be expected to occur in our waters. 2. Chordaria gracilis S. and G. Plate 84 Fronds 2.5-3.5 dm. high, with a single main percurrent axis 400-500/a diam. arising from a disk; primary branches numerous, short, scattered along the entire length of the main axis and alter- nately arranged on all sides, up to 4 cm. long, standing almost at right angles to the main axis; secondary branches 3-6 mm. long, standing 574 University of California Publications in Botany [Vol.8 at wide angles with the primary branches; color light brown; zoo- sporangia unknown; cortical filaments short, composed of 2-3 cells with a subspherical terminal cell. Growing in abundance on stones. West shore of Amaknak Island, Unalaska, Alaska. Setchell and Gardner, Phyc. Cont., 1924, p. 8. The slender percurrent axis bearing numerous short, almost equal, lateral branches over most of its length with short branchlets and all axes patent to one another, give the plants of the single collection available a very distinctive appearance. The plants are young and sterile, but the structure of the frond, the terminal subapical growth and the discoid holdfast indicate its affinities with Chord-aria. In the Algae of Northwest America (1903, p. 248) we placed this plant under Dictyosiphon foenicidaeeus (Huds.) Grew, an untenable posi- tion, since the subterminal growth and other details of structure plainly show its different nature. 3. Chordaria dissessa S. and G. Plate 41, fig. 57, and plate 75 Fronds attached by a small disk, 15-25 cm. high, up to 2 mm. thick, moderately and alternately branched, usually wider at the branching, but swollen in some parts and contracted in others, older parts at times fistulous, olive brown in color, composed of a medulla of longi- tudinal filaments giving rise to a cortex of short assimilating filaments ; filaments of the medulla composed of large, colorless, closely compact, thick-walled cells up to 120/x diam. in the central part of the frond, many times longer than the diameter, becoming much shorter and narrower toward the periphery ; cortical erect filaments cylindrical, simple or branched, not closely compact, 4-6^, diam., composed of 2-4 cells, terminal cell swollen and globular ; hairs unknown ; branches of each different order reduced in size, terminal ramuli long-attenuate and acute, angles mostly wide and rounded ; zoosporangia obovoid to ellipsoidal, 36-42^ long, 28-32/* broad ; gametangia unknown. Growing on eel grass in the middle and lower littoral belts. East Sound, Orcas Island, Washington. Setchell and Gardner, Phyc. Cont., 1924, p. 8. Castagnea divari- cata Setchell and Gardner, Alg. N.W. Amer., 1903, p. 249 (not C. (firaricata (Ag.) J. Ag.). The resemblance between Chardaria dissessa and C. divaricata (Ag.) Grev. is superficially close, but the former has its main axes less 1925] Setchell-Gardner: Melcmophyceae 575 regularly cylindrical, is less profusely branched, seems less gelatinous, with fewer or no hairs and has the assimilating cortical filaments shorter and of fewer cells, being 2-4 celled, while in C. divaricata they are 4_6 e ven more celled. The aspect of C. dissessa is that of a coarser, less abundantly branched, plant than is that of C. divaricata. 35. Analipus Kjellm. Thallus composed of an expanded, more or less deeply lobed, pseudoparenchymatous, sterile, basal portion attached to the sub- stratum by numerous, unicellular or multicellular, rhizoidal filaments, and giving rise above to simple, more or less cylindrical, solid or fistulous, fertile fronds; fertile fronds composed of a dense mass of pseudoparenchymatous cells with a few chromatophores giving rise on the surface to short, densely congested, cortical, assimilating fila- ments and to numerous zoosporangia ; gametangia unknown. Kjellman. Om Beringh. Alg., 1889, p. 48. The type and only known species of the genus is A. fusiformis from Bering Island. Analipus fusiformis Kjellm. Fertile fronds 2-6 cm. high, 1-3 mm. diam., one to several arising from the upper side of the same horizontal base; the sterile base pro- fusely and deeply lobed, terminal lobes rounded and blunt, attached by short, somewhat branched, monosiphonous hairs; cortical filaments of fertile fronds 3-5 cells long, elavate, with enlarged, rounded, terminal cell; zoosporangia arising from the base of the paraphyses, globose to ellipsoidal, 35/* long, 25/x broad (fide Kjellman), up to 65/x long. Growing on rocks. St. Paul Island, Bering Sea, Alaska. Kjellman, Om Beringh. Alg., 1889, p. 49, pi. 7, figs. 6-12 ; Setchell, Alg. Prib. Isl., 1899, p. 591 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 252. The material reported here seems to be almost identical with that found at Bering Island, the type locality, and reported by Kjellman. The zoosporangia are somewhat larger than the dimension given by Kjellman. some being as long as 65fi. Analipus seems to belong to the Chordariaceae as we have limited it, and we have been able to assure ourselves that the method of growth in length is subapical and, consequently, that the agreement with ;,7(; University of California Publications in Botany [Vol.8 Chordaria in family characteristics is close. In case Rupr&chtiella fUiformis (Rupr.) Yendo (Trav. Mus. l'Acad. Imp. Sci. St. Petersb., vol. 10, l!n.'!. p. 118) proves to be identical, generic-ally at least, with Anal i pus fusiform is Kjellm., Analipits must be removed to the Hetero- chordariaceae. 36. Gobia Reinke Frond filiform, more or less lubricous, simple or branched, hollow except in the very young stage, growth apical or subapical, composed of an inner tissue of loosely anastomosing colorless filaments giving rise to short, compact, vertical, assimilating filaments, to zoosporangia and to scattered colorless hairs; gametangia unknown. Reinke, Algenfl. westl. Ostsee, 1889a;, p. 65. The type of the genus, and the only other species known except the following, is Gobia. baltica, based upon material of Cladosiphon balticus Gobi (Brauntange finn. Meerbus., 1874, p. 12). The growth in length of the species of Gobia is said to be from an apical cell at first, but this soon ceases and intercalary growth sets in (cf. Reinke, 1889a, pp. 65, 66). If we are correct in referring the Mesogloia simplex Saunders to Reinke 's genus Gobia, with which it seems to agree in details of structure, we are probably also correct in interpreting the growth of the genus as being subapical in that the apical cell in very young plants and in very young branch initials very early cuts off a subapical cell which takes on the principal meristematic function. The same thing happens in Chordaria and presumably also in Myelophycus and in AnaUpus. This is very different from the situation in Dictyosiphon, where the terminal cell retains its meris- tematic function and activity as long as active growth in length con- tinues. Gobia, from this point of view, differs from the other genera of the Chordariaeeae in the loosely placed anastomosing filaments of its inner tissues. Gobia simplex (Saunders) S. and G. Plate 42, fig. 58, and plate 78b Fronds olive brown, 5-12 cm. high, 2-3.5 mm. diam., simple, cylin- drical or in part decidedly clavate, often arcuate, very blunt, solid in the juvenile stage, soon becoming hollow except at the very base of the apex, attached by a small disk ; central filaments loose, colorless, ana- stomosing freely, 25-35/* diam. ; cortical filaments straight, perpendi- cular to the main axis, the terminal segments composed of 2-4 cells. 1925] Setch ell-Gardner: MeUmophyceae 577 terminal cell pyriform, 30-35/a long:, 18-22/t diara., with large chroma- tophores; zoosporangia obovoid to ellipsoidal, 40-50/* long, 30-38/i broad ; hairs sparse, 4.5-5. 5/* diam. Growing on Heterochordaria abietina (Rupr.) S. and G. Extend- ing from Cook Inlet, Alaska, to Vancouver Island. Setchell and Gardner, Phyc. Cont., 1894, p. 12. Mesogloia simplex Saunders, Alg. Harriman Exp., 1901, p. 423, pi. 50, figs. 3, 4 ; Setchell and Gardner, Alg. N.W. Amer, 1903, p. 250. The measurements given in the diagnosis apply to the collection from Sitka. Saunders (loc. cit.) gives 25-35/a long and 15-25^ broad for the dimensions of the zoosporangia. Those in the collection of Policy and Butler in the Herbarium of the University of California (no. 99329) from Vancouver Island, have zoosporangia up to 80/x long. Otherwise they agree well with the Sitka plants. Small immature specimens of this plant were found growing on Heterochordaria abietina distributed by Miss J. E. Tilden (Amer. Alg., no. 348 sub Chordaria abietina.) . Gobia simplex differs from Gobia baltica in being simple and in its tendency to be somewhat enlarged above. Gobia simplex has always been found growing on Heterochordaria abietina. S. and G. family 17. COILODESMACEAE fam. nov. Fronds membranous, tubular or ligulate, growth subterminal at least in early stages, possible terminal at the beginning, composed of at least two tissues, inner composed of more or less elongated cells, becoming smaller outwards and passing into short anticlinal filaments firmly agglutinated into a cortical assimilating tissue; zoosporangia immersed in the cortical tissue but extending to, or near to, the sur- face ; gametangia unknown and gametophyte presumably microscopic. Coilodesmeae Kjellman, in Engler und Prantl, Die natiirl. Pflanzen fam., 1 Th., 2 Abt., 1893, p. 200. Stroemfelt (1886, p. 47) assigned his new genus Coilodesme to the Chordariaceae, as then understood, and as an intermediate type between the Chordarieae and the Scytosiphoneae. In vegetative struc- ture, Coilodesme is more highly differentiated than Chordaria, in that its cortical filaments have lost their distinctness and are combined into a definite tissue in which the zoosporangia are immersed. The method of growth has not been carefully or completely determined but seems (cf. Oltmanns, 1922, p. 62, fig. 354, 2) to be subterminal. 578 University of California Public at ions in Botany [Vol. 8 No species of CoUodesme has as yet been seen with gametangia and tliis suggests the extreme possibility of a microscopic gametophyte. This combination of characters leads us to suggest the new family Coilodesmaceae of the order Chordariales. Key to the Genera 1. Fronds solid, ligulate 38. Phaeostrophion (p. 585) 1. Fronds hollow, eomplanate or saccate 37. Coilodesme (p. 578) 37. Coilodesme Stroem. Fronds saccate, eomplanate from the beginning or more or less flattened in age, linear to ovate or obovate in outline, rounded above, tapering more or less at the base into a longer or shorter solid stipe, with small disk-shaped holdfast or with penetrating rhizoids epiphytic, parasitic or saxicolous, composed of two tissues, the inner consisting of large, irregular, thick-walled cells, with few chromatophores, the outer cortical portion consisting of a few to many layers of smaller color-bearing assimilating cells arranged more or less in radial rows; hairs absent ; unilocular zoosporangia developed within the cortex, scattered over the whole surface of the frond except the stipe and extreme basal portion ; paraphyses absent ; gametangia unknown. Stroemfelt, Meeresalg. Isl., 1886a, p. 173. The type species of the genus is C. bidlhgera, of Icelandic origin. The species of CoUodesme seem to divide themselves fairly defi- nitely into two groups, the one saxicolous, the other epiphytic. Of the epiphytic plants in the collections available to us, all are from the western coast of North America or the northeastern coast of Asia and seem to differ from one another essentially according to their host plant, only one host plant having more than one species associated with it. This is not surprising, since Rathbone (Journ. Linn. Soc, 1904, p. 671, pi. 24, figs. 5, 6) has shown the actual penetration of the host and our own investigations have demonstrated an intimate meta- bolic relation. The parasitism of all the epiphytic species of our coast is also striking in that the hosts are all members of the related genera, Halidrys, Cystoseira, and CystopkyUum, each of which seems to have one or more species of Coilodesme peculiar to it. Okamura (Icones Jap. Algae, vol. IV, no. 3, 1918, p. 55, pi. 144, figs. 10-13) figures and describes a Coilodesme Cystoseirae growing on Cystophyllum hakorfntense at Akkeslie, Japan. From the data presented, we are inclined to question the reference of this plant to C. Cystoseirae as interpreted by us, but careful comparison is necessary to determine L925] Setrhcll Gardner: Melcmophycew 579 exactly the nature of Ruprecht's Aspe.rococcus Cystoseircu and its relation to the Japanese plant on the one hand and the Alaskan plant on the other. Rupreeht states that the type of A. Cystosevrae is para- sitic on Cystoseira Lepidmm hut not on ('. thyrsigera. Key to the Species 1. Fronds complanate from the beginning, rigid, membrane up to .*^7o/u thick, epiphytic, stipe short 6. C. rigida (p. 584) 1. Fronds inflated at first, collapsing, especially when ruptured, membrane less than 250^ thick 2 2. Fronds ample, up to more than 1 m. long, and 1 dm. broad, epiphytic, stipe short 1. C. californica (p. 579) 2. Fronds smaller in all proportions 3 3. Fronds narrowly cylindrical, up to 8 mm. diam., and 8 dm. long, epiphytic, stipe short 5. C. Cystoseirae (p. 583) 3. Fronds more or less irregular, subcylindrical 4 4. Fronds finely corrugated, subsessile, epiphytic, stipe short 3. C. corrugata (p. 582) 4. Fronds undulate, decidedly stipitate 5 5. Fronds saccate, slightly undulate, saxicolous, stipe slender, elongated 2. C. bulligera (p. 581) 5. Fronds flattened, decidedly bullose, with ruffled margins, saxicolous, stipe slender, elongated 4. C. polygnampta (p. 582) 5. Fronds moderately bullose, epiphytic, stipe slender 7. C. sitchensis (p. 584) 1. Coilodesme californica (Rupr.) Kjellm. Plate 46, figs. 1, 2, and plate 86 Fronds flabby and fragile, smooth when young, more or less wrinkled, or bullate, and ruffled along the margins at maturity, inflated in the juvenile stage, soon collapsing because of ruptures, usually at the apex, chiefly cylindrical, rounded above, tapering rather abruptly at the base to a very short stipe 1-2 mm. long, variable in size from a few decimeters to over a meter long, up to 12 cm. wide, 175 to 215/x thick, composed of an inner tissue of 4-5 layers of large, colorless cells intermixed with smaller cells, and a cortical tissue of cells in anti- clinal rows, 2-4 cells in a row ; zoosporangia very irregular in form and size, 40-45/* long; color light brown, changing to greenish on exposure. Epiphytic on Cystoseira and Cystophyllum. Puget Sound region to southern California (San Pedro). Kjellman, Undersok. Slagtet Adenoeystis, 1889a., pp. 4-8, figs. 1-8 (excl. specimen) ; Saunders, Phyc. Mem., 1898, p. 160, pi. 29, figs. 1-3, Alg. Harriman Exp. 1901, p. 422; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 241; Collins, Mar. Alg. Vancouver Isl., 1913. p. 106; 580 University of California Publications in Botany [Vol.8 Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), Fasc. A, no. 1; Tilden, Amer. Alg. (Exsicc.) no. 354. Adenocystis (Lessonii var. ?) calif ornica Ruprecht, Tange Och., 1851, p. 291. Coilodesme amplissima, Setchell, Algae novae I, 1912, p. 232. The question as to the identity of the Adenocystis (Lessonii var.?) calif ornica of Ruprecht (loc. cit.) rests upon an imperfect description on the part of Ruprecht. The more definite statements of Kjellman and his illustrations seem to indicate that the plant intended was, at least, not ample, but in fact short and with a comparatively long slender stipe, dark brown, and an epiphyte. This was the inference of Setchell (1912, pp. 231, 232) although the herbarium specimens selected under that impression did not thoroughly correspond. Yendo (1913, pp. 279, 280) has stated that Kjellman 's Bering Island plant (loc, cit.) does not agree with Ruprecht 's type, which does agree with Tilden 's plant (no. 354, American Algae) from Vancouver Island which Setchell (lac. cit.) had referred to his Coilodesme amplissima. Yendo makes no statements as to characters or differences and his opinion is so little in agreement with the figures quoted by Ruprecht (viz., i>i Voy. Coquille, Bot. Atlas, 1826, pi. 11, fig. 2) that we hesitate to prefer his judgment to that of Kjellman. The important character separating the Coilodesme calif ornica of Kjellman and the C. amplis- sima of Setchell is the stipe, comparatively long and slender in Kjell- man 's plant and practically wanting in Setchell 's species. Ruprecht 's plant, however, was epiphytic on Cystoseira osmundacea, as is C. amplissima, while Kjellman 's plant is said to be epiphytic on larger algae. C. amplissima is widespread along the coast from Yakutat Bay, Alaska, to San Pedro, California. From the maze of uncertainty, it seems best to adopt Yendo 's statement as to the identity of Ruprecht 's plant and to regard C. amplissima as a synonym under C. calif ornica. Coilodesme calif ornica, as identified by Yendo, is an ample saccate species, the largest of the genus, with delicate walls, usually light yellowish brown in color, with extremely short stipe and parasitic on Cystoseira osmundacea and Cystophyllum geminatum. It extends from the Puget Sound region and possibly even from Yakutat Bay, Alaska (Saunders, 1898, p. 422), although the more northern refer- ences are subject to suspicion as to possible confusion with our C. sitchensis. Its southern limit is at San Pedro, California, so far as shown by available specimens. It is the only species as yet detected on the coasts of central and northern California where is situated Fort Ross, the designated type locality of the species. 1925] Setchell-Gardner: Melanopliyeeae 581 2. Coilodesme bulligera Stroem. Plate 45, figs. 77, 78 Fronds usually fasciculate, 5-12 cm. high, 15-22 mm. wide, 75-85/x thick, moderately firm, acuminate or rounded above, gradually nar- rowed below into a delicate stipe about a decimeter long, composed of 1-2 layers of large, thin-walled, colorless cells within and a cortical tissue of three layers of cells in vertical filaments or anticlinal rows; color dark brown, plants adhering firmly to paper on drying. Growing on rocks in the lower littoral belt, Shumagin Islands, Alaska, to Coos Bay, Oregon. .Stroemfelt, Meeresalg. Is!., 1886a,, p. 173, Om Algenveg. vid Isl. Kuster, 1886, p. 48, pi. 2, figs. 9-12 ; Saunders, Alg. Harriman Exp.. 1901. p. 422; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc), no. 923a; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 240 (in part). The saxicolous species of Coilodesme are found in the Arctic Ocean where it joins the North Atlantic (Iceland and Greenland) and in the northern portion of our own territory. Coos Bay, Oregon, is as far south as any has been found. The specimens collected by Gardner at Coos Bay agree well in habit and structure with some collected at Dixon Harbor, Alaska (somewhat northwest from Cape Spencer) by G. B. Kigg. The habit of both sets of plants is close to those col- lected and referred to C. calif ornica by Kjellman at Bering Island (cf. Kjellman, 1889a, figs. 1-3) but his were epiphytic and, conse- quently, to be suspected as being of some different species. The plants already mentioned agree also with the plant distributed by DeAlton Saunders from "Alaska" under 923a of the Phycotheca Boreali- Americana (at least in two copies in our possession) and with certain specimens collected by Gardner at Whidbey Island (no. 201). These plants all agree fairly well with the Stroemfelt figure of Coilodesme bidligera in being gradually and insensibly attenuated into a distinct slender, comparatively long stipe and in having the cortical layer of anticlinal rows of cells of three to four each and each terminating outwardly in a larger, dark brown, rounded cell. We are, consequently, referring all these to Stroemfelt 's species, but with some uncertainty. It is to be noted that 923b of the Phyc. Bor.-Amer. is referred by us to our C. polygnampta. ;,v L > University of California Publications in Botany [Vol. 8 3. Coilodesme corrugata S. and G. Plate 46, figs. 3, 4 Fronds very fragile and flaccid, very closely wrinkled all over the surface. 3-7 cm. Long, 8-14 nun. broad. 40-55/* thick, attached to the host by (penetrating?) rhizoidal filaments, rounded above, attenuated below into a short delicate stipe; eolor light brown; internal tissue composed of 2 layers of large, irregular, colorless cells and the cortical tissue composed of 2-3 layers of small, angular cells; zoosporangia irregular in form, usually broader than long. Epiphytic upon Cystoseira neglecta S. and G. In the upper sub- littoral belt. Santa Catalina Island, southern California, Setchell and Gardner, Phyc. Cont.. VII. 1024. p. 8. This is the smallest and most delicate species of the genus, and seems related to ('. amplissiina in general outline and structure. These characters, together with the delicate wrinkling ami its occurrence on a distinct and not common host, lead us to keep it distinct. Appar- ently it is quite limited in its distribution. 4. Coilodesme polygnampta S. and G. Plate Si 1 Fronds usually densely fasciculate, 1-4 dm. high, 1-5 cm. broad, moderately firm, complanate. at least at maturity, with definitely undulate and crisped margins, attached to rocks by a small, paren- chymatous, disk-shaped holdfast, rounded and blunt above, tapering more or less abruptly at the base into a definite, small, solid, cylin- drical stipe, dark reddish brown in color; histologically the interior tissue is composed of about 2-3 layers of colorless cells, the cortical tissue is composed of about 3 layers of small, assimilating cells in anticlinal rows. Growing on rocks in the lower littoral and upper sublittoral belts. Bering Sea. Setchell and Gardner, Phyc. Cont., VII, 1924. p. 9. Coilodesme buUigera Setchell and Gardner, Alg. N.W. Amer.. 1903. p. 240 (in part) ; Collins. Holden and Setchell. Phyc. Bor-Amer. (Exsicc), no. 923&. The plants referred from our Pacific Coast to Coilodesme buUigera Stroemf., do not all seem, on further study and careful comparison, to be the same as those of Iceland and Greenland. Our specimens 1925] Setchell Gardner: Melanopkyceae 583 from Bering Sea arc more ample, broader, tapering more abruptly into the slender stipe (at least in older specimens) than is represented in the figure of Stroemfelt (1886, pi. 2, fig. 9) or as compared with a specimen from Greenland distributed by Rosenvinge. In Stroem- felt's figure and in Rosenvinge's specimen, the cortex is made up of closely packed anticlinal rows of 3-4 cells each and each row ends outwardly in a somewhat Larger, dark brown, rounded cell. In the Bering Sea plant, the anticlinal rows are not recognizable, or are at most of only two cells, the outer deep colored cell usually sitting imme- diately upon the larger transparent outer cells of the inner layer. The typical form of our plant which we consider a distinct species is repre- sented by no. 923/; of the Phycotheca Boreali-Americana, and the type locality is Amaknak Island in the Bay of 1'nalaska. 5. Coilodesme Cystoseirae (Rupr.) S. and G. Fronds flaccid, irregularly cylindrical, blunt at the apices, taper- ing abruptly into a short stipe about 2 mm. long, 1-4 dm. (up to 8 dm.) high, 3-8 mm. diam., broad, usually inflated, composed of 2 layers of large, thin-walled, colorless cells on the interior and 2 layers of small, cuboidal, cortical cells ; zoosporangia ovoid, 15-20/x. long, ll-14,u wide. Epiphytic on Cystophyllwm . 248. Die/ iiosiphon Chordaria f. geiatinosa Setchell and Gardner, Alg. N.W. Amer.. 1903, p. 248 (not Stroemfelt). Scytosiphon iiip- puroides Lyngbye, Hydr. Dan., 1819, p. 63, pi. 14 B. D. liipj)uroi(lrs resembles D. fo< niciilaceus very closely, and is often confused with it. It is a coarser plant generally, is less profusely branched, each order of branches approximating the same size, and is less tapering. Collins (Rhodora, 1900, p. 164) states that this species frequently reaches a meter in length on the Atlantic coast. The mate- rial which we have referred to this species is scanty and of small size. Some of our plants may possibly be referred to the var. fragilis (Harv.) Kjellman. Order 6. LAMINAPJALES oltmanns Fronds of large size, solid or hollow, simple or branched, cylindrical to flattened, usually with three distinct regions, viz., (1) a holdfast varying from discoid to clusters of simple or branched hapteres, (2) a stipe cylindrical or more or less flattened and simple to dicho- tomously or irregularly branched, and (3) a flattened blade or blades, of usually three sets of tissues, inner of colorless, elongated, usually intertwined, hyphal cells, some with enlarged ends ("trumpet hy- phae"), intermediate of somewhat vertically elongated or nearly isodiametric cells with scattered chloroplasts (phaeoplasts) passing over into an outer layer of small cells usually deeply colored ; mucilage glands, or canals, present in some species ; zoosporangia on macroscopic sporophylls, borne in extended sori, accompanied by closely packed unicellular paraphyses having, usually, terminal hyaline appendages ; growth in length from a meristematic tissue intercalated between blade and stipe, or, in Chordaceae, just above the holdfast ; gametophyte microscopic, usually very much reduced, confervoid, bearing uni- locular gametangia ; sperms motile, biciliate ; eggs non motile, fertilized and germinating in position. Oltmanns, Morph, und Biol, der Algen (ed. 2), vol. 2, 1922, p. 121. The discoveries by Sauvageau (1915) and by Kylin (1916, 1918) of the gametophytes of several members of this group inaugurated a change in the whole attitude toward the classification of the Brown Algae and led to interpretations of certain peculiarities of the occur- rence of the reproductive bodies in certain groups. Following these 1925] Setchellr-Gwdner: Melanophycecu 591 indications to their seemingly logical outcome has led us to arrange the Brown Algae as we have, trusting to be justified, at least in large measure, when our knowledge of development within the group shall have been made more exact. The order of the Laminar iales, however, is clearly delimited by the intercalary region of growth in the complex plants belonging to it associated with the intermingling of the uni- locular zoosporangia with unicellular paraphyses in extensive elevated aori. The question as to the division of the order into families is not so clear as that of its cleavage from other orders. Two families have been recognized and there is great temptation to increase the number. There certainly exists a certain tendency to group genera into some sort of larger divisions, but the difficulty lies in finding sharp lines of cleavage. The best distinctions lie according to the differentiations of the frond arising at the transition place between the blade and the frond and the relation to this and the fertile areas. The alternatives Beem to be these: either to retain all the genera under one large and varied family (possibly excepting the Chordaceae) divided into num- erous tribes, or to divide them between numerous families of unequal value and of often indistinct cleavage. We have separated the Lami- nariaceae into five families, one of which Haligeniaceae (including Haligenia and Sasccorhiza) is not represented in our territory. Key to the Families 1. Stipe absent or not differentiated, paraphyses without hyaline appendages 19. Chordaceae (p. 591) 1. Stipe distinct, at least when young, paraphyses with hyaline appendages 2 2. Transition place neither giving rise to splitting nor to outgrowths 20. Laminariaceae (p. 593) 2. Transition place not as above 3 3. Splitting arising at transition place or within its influence 21. Lessoniaceae (p. 621) 3. Outgrowths arising at transition place or within its influence 22. Alariaceae (p. 633) family 19. CHOEDACEAE dumort. Fronds cylindrical, hollow, interrupted by diaphragms, with a discoid holdfast, growth in length basal (situated just above the hold- fast), innermost cells elongated, supporting hyphal filaments united to form the diaphragms ; outermost layer a single row of slightly elon- gated cells with colorless hairs and with or without exserted colored filaments ; sori extended, paraphyses destitute of colorless appendages ; gametophyte microscopic as in the order. 592 University of California Publications in Botany [Vol.8 Dumortier, Comm. Bot., 1822, p. 72 (fide Pfeiffer) ; Kylin, Stud. iiber Algenfl. der Schwed. Westkiiste, 1907, p. 97. The single genus Chorda, clearly a member of the Laminariales, but differing from all the other genera of the order in its hollow cylindrical frond, is taken to represent a distinct family. 40. Chorda Stackh. Fronds arising from a small disk-shaped holdfast, unbranched, cylindrical, solid below, hollow above, the cavity separated by cross- partitions into many chambers, color olive brown, growth intercalary near the base; median tissue of the hollow portion composed of longi- tudinally elongated cells, hexagonal in cross-section, firmly united, lined wath delicate, loose filaments which unite to form the cross-parti- tions at intervals; the outside of the median tissue is clothed with unicellular paraphyses, hairs and zoosporangia, ; reproduction asexual by unilocular zoosporangia. Stackhouse, Physiol. Observ., in Ner. Brit., 1797, p. xvi. The type of the genus is the Fucus Filum of Linnaeus (Sp. PI., 1753, p. 1162). The type locality is unknown. Chorda Filum (L.) Lamour. Fronds gregarious, several at times arising from the same confluent base, cylindrical, sometimes spirally twisted, slightly tapering at the base and at the apex, from 3 dm. to 4 m. long, 2-5 mm. diam., lubricous, w r hen young clothed with a dense growth of hyaline or yellowish colored hairs; paraphyses unicellular, densely crowded, clavate, slightly longer than the zoosporangia and nearly obscuring them above ; unilocular zoosporangia oblong to ellipsoidal, 30-50ju long, 10-15//, wide; plurilocular gametangia 35-40/x wide. Growing on stones in the upper sublittoral belt. From Port Clarence, Alaska, to the Straits of Juan de Fuca. Lamouroux, Essai, 1813, p. 26; Saunders, Alg. Harriman Exp., 1901, p. 424; Setchell and Gardner, Alg. N.W. Amer.. 1903, p. 254; Reinke, Atlas, 1889, p. 35, pis. 26, 27, 28. Fucus Filum Linnaeus, Sp. Plant., 1753, p. 1162. There is little to be said under Chorda except to note that it is a plant of the colder waters and of which we have seen specimens from our coast collected at Port Clarence, Safety Harbor, near Cape Nome, 1925 1 Setchell-Ga/rdner: Mvlanophyceae 593 St. Michaels, and Golofin Bay on the Alaskan coast. It is credited by Postels and Rnprecht (1840, p. 19) to Sitka and Unalaska, by Saunders (1901, p. 424) to Popof Island and Prince William Sound, and by MacMillan (1902, p. 219) to the Straits of Juan de Fuca. The last locality is not verified by any specimens, so far as we know. Such specimens as we have seen do not differ from those of the North Atlantic. family 20. LAMINARIACEAE reichenb. Fronds simple, composed of holdfast, stipe and blade, at least in the earlier stages, never with true dichotomous branching of the stipe, adventitious or false branching arising in a few cases from the blade, composed of the usual three tissues; mucilage glands and passages present or absent; blades destitute of cryptostomata (or bunches of hairs) ; sori extended, always on the blade, paraphyses with hyaline appendages. Reichenbach, Conspect. Reg. Veg., 1828, p. 29 (fide Pfeiffer). We have adopted the family of the Laminariaceae in restricted sense as including those members of the Laminariales having the prin- cipal meristematic region situated at the summit of a more or less well developed stipe, at the base of a flattened blade, and initiating neither splitting nor outgrowths, of or from, either the stipe or the blade. In the simplest genera (e.g., Laminaria) the transition place is perfectly plane, but in some genera folds (e.g., Pleurophycus and Cijmathaere), in others ribs (e.g., Costaria and Agarum) are developed, while in some genera (Costaria, Agarum, and ThalassiophyUum) bullosities and resulting perforations are the final results of its activity. In certain species the center of the blade wears away to the transition place which extends and growing on, produces two false branches each with half a blade attached {Hedophyllum subsessile and ThalassiophyUum) . It has seemed best to group the various genera into tribes according to these various methods of behavior. Key to the Tribes 1. Blades without perforations, longitudinal ribs or folds, occasionally bullose.... 2 1. Blades with perforations, longitudinal ribs or folds 3 2. Stipe persistent 1. Laminarieae (p. 594) 2. Stipe early disappearing 4. Hedophylleae (p. 616) 3. Blades with longitudinal folds 2. Cymathaereae (p. 606) 3. Blades with longitudinal ribs or perforations 3. Agareae (p. 609) f>!)4 University of California Publications in Botany [Vol.8 tribe 1. LAMINARIEAE bory (lim. mut.) Fronds with *well developed stipes and flattened blades; hapteres sometimes elongated and producing new fronds; blades plane, undi- vided or splitting from above downwards, at times bullose, but without perforations, longitudinal ribs or folds. Bory de Saint Vincent, in Diet. Class., vol. 1, 1822, and vol. 9, 1826, p. 187 (as family) ; De-Toni, Syll. Alg., vol. 3, 1895, p. 317 (in part) ; Setchell, Kelps of the U. S. and Alaska, 1912*, p. 146. The tribe of the Laminarieae contains the single large genus, Laminaria. It may be looked upon as containing the simplest, or, certainly, the least modified of the members of the family and as the type whence divergences have arisen or taken their departure. 41. Laminaria Lamour. (in part) Fronds differentiated into three distinct parts, a basal holdfast, composed either of a solid disk of more or less branched hapteres, a stipe of greater or less length, and a blade ; both the stipe and the blade may or may not have mucilage ducts; blade simple or more or less deeply lacerated into few to many segments, plane or with intra - marginal bullae and marginal ruffles ; growth intercalary at the base of the blade; reproduction asexual, by zoospores borne in unilocular zoosporangia among unicellular paraphyses in extensive sori nearly covering both surfaces of the blade of the macroscopic plant, and sexual, oogamous, on microscopic plants. Lamouroux, Essai, 1813, p. 20. The Laminaria of Lamouroux (loc. eit.) included plants now refer- red to several genera and well distributed through the order Lamin- ariales. The first three species, however, are of the genus Laminaria as gradually restricted by later authors. The Lamouroux name is antedated by the Laminarius of Roussel (1806), by the Saccharina of Stackhouse (1809) and the Phyeodendron of Olafsen and Povelsen (1772). Laminaria has, however, received the sanction of the Inter- national Botanical Congress of Brussels (1910, see Briquet, 1912, p. 76). While most of the species of Laminaria are readily separated into those whose blade is digitate (i.e., split deeply into segments) and those whose blade is entire, there are a few species seemingly present- ing both conditions, e.g., L. cuneifolia of our coasts. 1925] Setchell-Gardner: Melamophyceae 595 Key to the Species 1. Blade entire or slightly split at the outer end 2 1. Blade split more or less deeply into several narrow segments 8 2. Fronds, except when young, arising from creeping "rhizomes" 3 2. Fronds without creeping "rhizomes" 4 3. Mucilage ducts absent from stipe 3. L. longipes (p. 597) 3. Mucilage ducts present in both stipe and blade 4. L. Sinclairii (p. 598) 4. Holdfast a solid parenchymatous disk 6. L. personata (p. 599) 4. Holdfast of branched hapteres 5 5. Blade densely bullate all over 5. L. Farlowii (p. 599) 5. Bullae in two rows within the margin or scattered 6 6. Stipe long, complanate, except at base, very dark brown 2. L. complanata (p. 596) 6. Stipe short, terete, color yellowish brown 7 7. Mucilage ducts in the blade only 1. L. saccharina (p. 595) 7. Mucilage ducts in both blade and stipe 7. L. cuneifolia (p. 600) 5. Holdfast a solid parenchymatous disk 8. L. ephemera (p. 603) 8. Holdfast composed of stout branched hapteres 9 9. Fronds plane, stipe long, up to 1 m 10 9. Fronds more or less bullate, stipe relatively short 7. L. cuneifolia (p. 600) 10. Stipe cylindrical to near apex 11 10. Stipe compressed from just above holdfast 11. L. platymeris (p. 605) 11. Mucilage ducts of stipe in a circle just beneath the cortex, stipe slender 9. L. dentigera (p. 604) 11. Mucilage ducts of stipe deep seated, stipe rigid 10. L. Andersonii (p. 605) 1. Laminaria saccharina (L.) Lamour. Holdfast of numerous, branching, more or less rigid hapteres often extending for 5-8 cm. up the stipe which is very variable in length, from 5-50 cm., cylindrical, flattening above into the blade, 6-9 mm. diam., without mucilage ducts ; blade undivided, usually coriaceous or membranaceous when growing in quiet waters, plane, undulate, or with two distinct rows of large bullae along the margin, ovate to ovate- lanceolate, often with cuneate base, with mucilage ducts; color rich brown. Growing on rocks in the upper sublittoral belt. Reported from various localities from the Alaskan peninsula to Coos Bay, central coast of Oregon. Lamouroux, Essai, 1813, p. 22; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 261 ; Setchell, Kelps of the U. S. and Alaska, 1912a, p. 149. Fucus saccharmus Linnaeus, Sp. PI., 1753, p. 1161. All attempts to separate this widespread species of the cooler waters of the northern hemisphere have been unsatisfactory. The type has been assumed to be the plant with mucilage ducts in the blade, but lacking them in the stipe. This seems to be a condition with all the 596 University of California Publications in Botany L VoL -8 plants we have referred here. We have not met with any forms referrable to L. Ayardhii, i.e., entirely destitute of mucilage ducts. On our coast at least two forms of this species may easily be distinguished : Laminaria saccharina f. linearis J. Ag. Holdfast composed of an abundance of relatively small branched hapteres extending along the stipe up to 8 cm. above the base, stipe slender and short, blade with cuneate base, 12-18 cm. broad, 2.5-3.5 m., or more, long, with two distinct rows of bullae extending the entire length of the blade just within the margin. Growing on rocks in localities subjected to the action of the water, in the upper sublittoral belt. From Unga, Alaska, to Puget Sound, Washington. J. Agardh, De Lamin., 1867, p. 12 ; Saunders, Alg. Harriman Exp., 1901, p. 429, "forma b."; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 261 ; Collins, Holden and Setchell, Phyc. Bor.-Amer. (Exsicc.), no. LXXXVIII. Laminaria saccharina f. membranacea J. Ag. Hapteres strong, stipe longer than in the preceding, frond ample, membranaceous, often of delicate consistency, not holding together when lifted from the water, more or less undulate or bullate; large fruiting specimens at Sitka, Alaska, are stiff and rigid, not readily torn. Growing on rocks, logs, floats, etc., in the upper sublittoral belt. From Alaska southward to Coos Bay, Oregon. J. Agardh, De Lamin., 1867, p. 13 ; Setchell and Gardner, Alg. N.W. Amer., 1903, p. 261 ; Saunders, Alg. Harriman Exp., 1901, p. 429, "forma a." Laminaria saccharina Tilden, Amer. Alg. (Exsicc), no. 240. 2. Laminara complanata (S. and G.) Setchell Plate 57 Holdfast of few, rigid, slightly branched hapteres ; stipe 3-5 dm. long, terete below, much flattened above, varying from 8-12 mm. diam. below, to 20-25 mm. wide and 2-4 mm. thick above, without mucilage ducts; blade 80-100 cm. long, 40-50 cm. wide just above the base, in many specimens short and nearly circular in outline, amply ruffled, L925] Setrhell-Gardner: Mclanophyc