THE STUDY OF SEAWEEDS AN a. INTRODUCTION TO THE STUDY ^ OF A/ C SEAWEEDS BY GEORGE MURRAY, F.R.S.E, RL.S. KEEPER OF THE DEPARTMENT OF BOTANY, BRITISH MUSEUM WITH EIGHT COLOURED PLATES AND EIGHTY-EIGHT OTHER ILLUSTRATIONS Uontion MACMILLAN AND CO. AND NEW YORK 1895 The Right of Translation and Reproduction is Reserved RICHARD CLAY AND SONS, LIMITED, LONDON AND BL'NGAY. PREFACE SINCE the last introduction to the study of sea- weeds was written, many years ago, the aspect of the whole subject has been completely changed by the progress of research. I have attempted in the following pages to keep the rule of describing only what I have personally verified by examination or by inspection of the original account, and this has been possible in nearly every case. It would have been more in accordance with usage to begin with the sub-class Rhodopliycecc, but I have permitted considerations of convenience to prevail. The Rhodophyccce present so many diffi- culties, to be understood only after the study of other groups, that I have chosen Phwophyccm, with its familiar forms of sea- wracks and tangles, for the first sub-class. The Chlorophycem and Diatomacccc follow naturally. The Rhodophycem next make a vi PREFACE series by themselves, and finally come the simple Cyanophyccce. The account of the Rhodophycccv is based on the scattered papers of Schmitz (p. 37), who by utilising his own researches and the splendid investigations of Thuret and Bornet has almost wholly altered the classification of this sub-class. A mere outline of his results has been published with the effect of destroy- ing the earlier classification and of incompletely establishing his own. The full justification of his system will, however, be published soon, having been left by him almost finished at his lamented early death. I have thought it better to give here from the fragments of his work accessible to me, and from other sources, an account of the sub-class under his system, rather than revert to the old classification. I have to thank the Council of the Linnean Society, the editors of the Annals of Botany and Journal of Botany, and Messrs. Dulau and Co., publishers of the Phycological Memoirs, for their kindly granted permission to reproduce copies of my own and other figures. I am also indebted to M. Bornet, Prof. Reinke, Graf zu Solms-Laubach, Dr. John Murray, Mr. Edward Trewendt of Breslau, Dr. Hauptfleisch, Dr. Kuckuck, Mr. Richards, and Miss N. Karsakoff for their courtesy and kindness in granting me permission to copy some of their figures; to Mr. Antony Gepp for assistance in proof-reading and for the use of two photographs ; to Mr. George Brebner for kindly drawing for this book Figs. 10 and 79, and to Mr. G. West for Fig. 87. Above all, I am indebted to Miss E. S. Barton and Miss A. Lorrain Smith for many drawings made specially for this book. GEORGE MURRAY. TABLE OF CONTENTS PACK INTRODUCTION 1 LITERATURE 34 SUB-CLASS I. 39 FUCACE.E 40 CUTLERIACE.E 56 DICTYOTACE.E 60 TlLOPTERIDACE^ 66 SPLACHNIDIACEJS 70 LAMINARIACEA: 75 SPOROCHNACEA; 86 CHORDARIACEJE 90 ELACHISTACEA: 93 DlCTYOSIPHONACEvK 98 DESMARESTIACE^K 99 STRIARIACE^: 101 ENCXELIACE.E 104 RALFSIACE^E 108 SPHACELARIACE.E Ill CHORISTOCARPACE^J 115 ECTOCARPACE^ 116 60983 x CONTENTS SUB-CLASS II. PAGE CHLOROPHYCE.E 120 CAULERPACE.E 121 VAUCHERIACE.E 127 CODIACE^E 132 UDOTEACE^E 137 DASYCLADACE^S 145 Acetabularieai . . • 145 Dasycladeai 151 VALONIACE^E 156 CLADOPHORACEJE 165 ULOTRICHACE.E 170 ULVACE^S , . 174 PROTOCOCCACE^J 177 PLEUROCOCCACE^EJ 179 PERIDINIE^E 181 COCCOSPHERES AND RlIABDOSPHERES 185 SUB-CLASS III. DIATOMACE.E 188 SUB-CLASS IV. RHODOPHYCE.E OR FLORIDE^: 200 NEMALIONACEJE 207 Helminlhodadiete 207 Chtetangiece 210 Gelidiea 214 GlGARTINACE^K 216 RHODYMENIACL^E 222 Sphwrococcew 223 Rhodymenieai 224 CONTENTS xi PAGE Delesseriete . 230 SonnemaisoniecK 232 Rhodomeleai 233 Ceramiece 235 CRYPTONEMIACE^E 238 Gloiosiphoniece 238 GrateloupiecB 238 Dumontiece 238 Nemastomece 238 Rhizophyllidece 239 Squamariece 240 CorallinecK 241 BANGIACE^E 246 SUB-CLASS V. CYANOPHYCE^E 249 NOSTOCACE^J 251 Heterocystew 252 Homocystece 255 CHROOCOCCACE^E , 258 ISDEX . 265 LIST OF COLOUEED PLATES PLATE I Tofacepaye 39 1. PELVETIA CANALICULATA 2. HALIDRYS SILIQUOSA 3. CYSTOSEIRA ERICOIDES 4. CUTLERIA MULTIFIDA PLATE II „ 60 1. PADINA PAVONIA 2. CHORDA FILUM 3. ASPEROCOCCUS ECHINATUS 4. SPOROCHNUS PEDUNCULATUS PLATE III „ 120 1. CODIUM TOMENTOSUM 2. CODIUM BURSA 3. HALICYSTIS OVALIS 4. BRYOPSIS PLUMOSA PLATE IV „ 165 1. ULVA LACTUCA 2. ENTEROMORPHA INTESTINALIS 3. CLADOPHORA RUPESTRIS LIST OF COLOURED PLATES xiii PLATE V To face page 207 1. PORPHYRA LACINIATA 2. SCINAIA FURCELLATA 3. NEMALION MULTIFIDUM 4. NACCARIA WIGGHII PLATE VI „ 216 1. PHYLLOPHORA RUBENS 2. CATENELLA OPUNTIA 3. CYSTOCLONIUM PURPURASCENS 4. GlGARTINA MAMILLOSA 5. CHONDRUS CRISPUS PLATE VII „ 222 1. CERAMIUM DIAPHANUM 2. RHODYMENIA PALMATA 3. DELESSERIA SANGUINEA 4. POLYSIPHONIA BRODIAEI 5. LOMENTARIA ARTICULATA PLATE VIII „ 238 1. POLYIDES ROTUNDUS 2. MELOBESIA MEMBRANACEA 3. DCMONTIA FILIFORMIS 4. CORALLINA OFFICINALIS LIST OF ILLUSTKATIONS IN TEXT FIG. PAGE 1. Reinke's Dredge 28 2. -Fucus plant with organs of reproduction 41 3. Coccophora Langsdorfii 43 4. Himanthalia lorea 46 5. Notheia anomala 47 6. Conceptacles of Sarcophycus potatorum 50 7. Turbinaria conoides and T. Murray ana with growing- points 53 8. Cutleria multijida, sori 56 9. Didyota dichotoma, sori 62 10. Scaphospora speciosa, and Haplospora globosa 67 11. Splac hnidium rugosum 71 12. Alaria esculenta and cryptostoma of Saccorhiza ... 75 13. Laminaria digitata 77 14. Agarum Turneri 77 15. Postelsia palmcpformis 78 16. Macrocystis pyrifera 80 17. Macrocystis and Postelsia, sori 83 18. Adenocystis Lessonii 84 19. Spermatochmis paradoxus 88 20. Stilophora rhizodes, growing point 89 21. Leathesia difformis, sporangia 91 22. Chordaria divaricata, sporangia 92 23. Myriotrichia clavceformis 96 24. Desmarestia aculeata, secondary thickening 100 25. Kjellmania sorifera, sporangia 103 LIST OF ILLUSTRATIONS xv FIG. 1'AGE 26. Stictyosiphon Decaisneii, section 103 27. HydroclatJirus, sporangia and cryptostomata 106 28. Sphacelaria, sporangia, &c 112 29. Ectocarpus confervoides 118 30. Caulerpa phyllaphlaston 122 31. Caiderpa cactoides 123 32. Caiderpa ligidata and G. Holmesiana 125 33. Caiderpa Carruthersii 126 34. Vaucheria synandra 129 35. Codium tomentosum, sporangium 132 36. Bryopsis Flanagani • 134 37. Avrainvillea, structure 138 38. Penicilhis capitatus 140 39. Rhif)ocephalus Phoenix 141 40. Udotea Pavonia 142 41. Halimeda monilis 143 42. Acicularia Schenckii 145 43. Acetabidaria and Acicularia, structure 146 44. Acetabidaria crenulata 149 45. Neomeris and Bornetella 153 46. Valonia and Halicystis 158 47. Dictyosphceria, structure 160 48. Strnvea macrophylla, S. plumosa, and S. tennis .... 162 49. Boodlea coacta 164 50. Gomontia polyrhiza 167 51. Urospora penicilliformis .-.. 168 52. Chcetomorpha Darwini 169 53. Botttocoleon piliferum ....171 54. Letterstedtia insignis 175 55. Halosphwra viridis 178 56. Ceratium Tripos 183 57. Pyrocystis noctiluca and P. fasiformis 184 58. Rhabdosphere and Coccosphere 186 59. Pinnularia viridis, and diagram of successive divisions of a Diatom 193 xvi LIST OF ILLUSTRATIONS FIG. PAGE 60. Pelagic Diatoms (Chcetoceras, Planktoniella, and Bacte- riastrum) 196 61. Tetraspores 206 62. Chantransia corymbifera 208 63. Chcetangium ornatum, and procarp of Scinaia fnrcellata 211 64. Choreocolax Polysiphonice and C. albus 212 65. Carpogonium of Naccaria hypnoides, and plant of Geli- dium corneum 215 66. Stenogramme interrupts 217 67. Callophyllis obtusifolia, cystocarp 218 68. Eticheuma spinosum 220 69. Catenella opuntia, carpogonium 221 70. Phacelocarpus Labillardierii and cystocarp of Graci- laria confervoides 223 71. Chylodadia kaliformis, carpogonium and young cysto- carp 225 72. Plocamium corallorhiza 227 73. Ckampia parvida, germination of carpospore 229 74. Claudea elegans 231 75. Martensia elegans 232 76. Cliftoncea pectinata 234 77. Dasya elegans, diagram of procarp for Hhodomelece and procarp of Polysiphonia 234 78. Ceramium diaphanum and procarp of CallitJiamnion . 236 79. Gloiosiphonia capillaris, development of cystocarp . . 237 80. Gloiosiphonia capillaris, procarp 238 81. Dudresnaya, fertilisation 239 82. Polyides rotundus, fertilisation 240 83. Lithothamnion polymorphum 242 84. Corallina mediterranea, antheridium 243 85. C. mediterranea, tetraspores and cystocarp 245 86. Rimdaria hospita and Calothrix pidvinata 253 87. Dermocarpa, sporangia, &c f. 259 88. Hyella ccespitosa 260 SEAWEEDS INTRODUCTION THE study of seaweeds is of very modern origin, and nothing beyond casual recognition of their existence is to be found in the literature and memorials of early times. The Greeks have left us engraved figures of Gorgons whose heads were decorated with seaweeds ; there is but one mention of them in the Bible, when Jonah exclaims, " The depths closed me round about, the weeds were wrapped about my head '' ; and the re- ferences in Latin literature, even that of the poets, such as the " Alga projecta vilior " of Virgil and the " inutilis Alga " of Horace are merely contemptuous. While other plants received notice and were the subjects of study in these early times and during the middle ages, the flora of the sea remained within its confines — a hortus indusus within a barrier that still jealously hides much from our knowledge. In Sir Hans Sloane's great herbaria of many travellers and collectors, preserved in the £ B 2 SEAWEEDS British Museum, there is the earliest authentic evidence of the collecting of seaweeds, the beginning of the study; and the foundation of its literature was laid by later systematic writers, including Linnaeus. It was only to be expected that many marine animals, such as Zoophytes, which resemble seaweeds frequently in outward form, should have been indiscriminately classed with them by these writers, and it was not until the present century, when our knowledge of minute structure had advanced, that a strict division became possible between the stony coralline Algse and similarly encrusted animals. Gmelin's Historia Fucorum (1768) and Esper's Abbildungcn der Tange (179*7) were the first noteworthy efforts to gather within a book devoted to the study of AlgaB all that was then known, and as the result of the stimulus so imparted to research, the first years of the present century witnessed greater activity and progress in the accumulation of knowledge of the forms of sea- weeds and their classification. Lamouroux pub- lished his Dissertations sur plusieurs especes de Fucus in the year " XIII "( = 1805) of the new era of the French Revolution, and a few years later there was begun the best of all the early books, Dawson Turner's Fuci (1808-1819), which not only cleared up many of the difficulties of preceding writers, but presented a large body of new facts acquired from the study of specimens brought home by Robert Brown and other great botanists and travellers of that time. Perhaps the last of those who may be called the pioneers of Phycology was Lyngbye, whose Tentamen INTRODUCTION 3 Hydrophytologicc Daniccc was published in 1819 and dedicated to " Frederic VI., King of Denmark, and of the Goths and Vandals." Just as Linnaeus was the last of the older naturalists as well as the first of the new, so the elder Agardh makes a link for us in the history of Phycology. With his greater son, happily still alive, he laid the foundations of the present system of classifying seaweeds, while their fellow-countryman Fries was performing a like service for the study of Fungi, continuing the work of Linnaeus in the spirit of Linnseus and in the land of the great naturalist. Germany contributed Kiitzing to the group of great systematic writers of the same period, and though his work is characterised more by his extraordinary industry than by any new departure of system or method, it has greatly influenced the study by increasing the sum of knowledge and the facilities of reference. Our countrymen Harvey and Greville achieved yet greater advances. The former, by his travels and his genius as a collector, describer, and depictor of marine Algae, surpassed all others in this field ; while the latter, in addition to his great services to other departments of crypto- gamic botany, has left observations of the minute structure of seaweeds that no subsequent research has shaken, and has done much towards establishing a natural system of classification. Thuret, and his fellow-worker Bornet, brought to its present state of development the methods of minute study of structure and development that only need wider ap- plication in the future to ensure the advancement of Phycology. B 2 4 SEAWEEDS The first observation commonly made by a student of seaweeds is of the variation of their colours. The green hue that prevails throughout land vegetation, except in the colours of flowers and the bark of trees, is varied in the case of seaweeds with olive- brown, and red forms. An artificial classification of them according to their colours leads to the striking result that it nearly coincides with the natural classi- fication of them according to their structure and development. Such an artificial classification be- came firmly established, and has left its mark on the names of the natural primary divisions or sub-classes of Algse, viz. the Rhodophycew or Red Seaweeds ; the Phceophycece, or olive-brown; the CUorophycece or green; and the Cyanophyceas or blue-green. A simple experiment proves that fundamentally they are all green, and that the red colouring matter phycoM-ythrine, the brown phycophceine, the yellowish- brown phycoxanthine, and the blue pJiycocyanine are each something added to the chlorophyll or leaf-green that characterises vegetation in general, and by virtue of which plants form the organic substances necessary for their nutrition. These additional colour- ing matters can be extracted by fresh water, leaving the previously red, olive, &c. plants green, and they differ from the green colour in this respect, since it is insoluble in water. Though there occur excep- tionally a few red forms, numerous blue-green, and (in the diatoms only) many brown forms in fresh- waters, there still remains the broad fact that these colouring matters are characteristic of seaweeds, and it is in the conditions of plant-life in the waters of INTRODUCTION 5 the sea that an explanation of their nature must be sought. It has been found that the colours of sea- weeds are more or less indicative of their range in depth in the sea, and, allowing for numerous excep- tions, that there is a zonal distribution of Algae according to their colours. The uniformity of this distribution is disturbed by the fact that the con- ditions are not equal for all in the face of the deter mining influence, as will presently be made plain As a general rule the inshore seaweeds near high- water mark are green in colour like land vegetation and lower down between tide-marks there is a belt 'of olive forms sheltering red plants beneath them. Where rocks overhang the bottom, and in small pools these red forms also occur at this level. At extreme low-water mark, and beyond it, are found the brown tangles sheltering red forms again, while at the lowest depths of plant-life in the sea the red forms occur without shelter. Between 20 to 50 fathoms seaweeds become more and more rare, while below that depth their occurrence is exceptional. That the main influence determining this regulation of pigment is the nature of the supply of sunlight, necessary to the action of chlorophyll in the work of nutrition, is apparent from the following facts. The interception of sunlight by sea- water brings about a state of total darkness at 700 fathoms, probably less, and though seaweeds do not penetrate to a depth approaching this limit of light, a further consideration will account for their failure. Not only is the quantity of sun- light reduced by its passage through the water, but its quality is affected, as spectroscopic investigation 6 SEAWEEDS has shown. It is precisely those rays that are most efficient in the work of assimilation by plants that are first intercepted, and only the blue and green rays travel to greater depths. It may be taken, then, that the red, brown, and yellow colouring matters, added to the fundamental green, are adaptations to the supply of sunlight. Whether they act in the direction of heightening the susceptibility of chlorophyll to a diminished supply of the useful rays, or as a protection against a relative excess of the blue rays, has not been settled experiment- ally, but the balance of probability is in favour of the latter theory, since it has been discovered that certain pigments in other plants act as a shield against illumination of this character. A microscopic green Alga, Halosphcera viridis, has been obtained from the great depths beyond the reach of sunlight, and the speculation has been hazarded that the lumin- osity of animals inhabiting those regions might in its case be an efficient substitute for sunlight, but the idea is wholly unsupported by experimental evidence. The explanation that the plants in question were swept there by currents of submerged waters is much more in accordance with oceanographical facts. The fact that colour, which affords a character of notorious instability in determining claims merely to specific rank among land plants, should be found associated in the Algae with characters of more than ordinal importance (though not constituting such characters by itself) is not so puzzling when it is remembered that it plays here a role of vital import- ance in the matter of nutrition. INTRODUCTION 7 As light is the factor that determines the zonal distribution of seaweeds, and thus influences their local habitats, so temperature is the leading in- fluence, among others of minor potency, that affects their geographical distribution. They inhabit a medium of stable temperature, in which they are not called upon to adjust themselves to any great periodical or fortuitous changes, varying little from day to night and from season to season. On this ground alone they might be presumed to be pecu- liarly sensitive to change of temperature, and experi- ments in the culture of seaweeds in aquaria show that a nice regulation of temperature is necessary to success. Comparisons of the marine flora of areas of different temperature confirm this view though they do not exclude other possibilities, since the general character of such floras is modified by other in- fluences— the nature of the bottom to some extent, the degree of salinity of the water, the presence, absence and amount of the tides, &c. Marine vegetation, like fresh-water vegetation, is removed from the influence of relative humidity which determines frequently the character of a land flora, but on the other hand it is subject to control by factors such as relative salinity and the like. The contour of the earth's surface, which brings about the existence of alpine floras for example, and frequently affords many climates at the same latitude on land, has no corresponding influence on the marine flora, since conditions of illumination check range in depth. On the whole, temperature may be said to be left more to itself as a determining influence of the character of marine 8 SEAWEEDS floras than of land floras, and when its operation is tested by a survey of the pelagic vegetation, this view is much enforced by the result. Pelagic free floating Algae removed from the influence of coast and river waters, rise and fall of tides, nature of bottom, &c., and left to the nearly exclusive operation of temperature in stamping their char- acter, are found to respond to this influence and to vary with areas of temperature in the sea. As an extreme term in resistance to adverse conditions of temperature and supply of light, it is interesting to note Kjellman's observation of the normal growth and fructification of Algae during the dark arctic night and at a mean temperature of— 1°C. Obser- vations of any notably high temperatures resisted in the sea arc of course of small physiological interest compared with the temperatures resisted by fresh- water Algae inhabiting hot springs. In the culture of seaweeds in aquaria it has been found that forms from deep water are peculiarly susceptible to rise of temperature and undue illumin- ation, so much so that merely for their transport it is necessary to choose a cloudy day, especially if in summer, and the use of ice is almost always advis- able. A cool chamber from which direct sunlight is excluded is a condition of success in the culture of most seaweeds. One way in which these plants may be killed by too much care is in the attention paid to aeration of the water. Very little is necessary, since the air so introduced has been found to carry off too much CO2. A sudden change of water is also mis- chievous ; and added water (whether fresh- water to INTRODUCTION 9 make up for evaporation, or salt-water) should be supplied drop by drop. One of the best ways of cultivating seaweeds is by suspending them in baskets in the sea at proper depths from anchored buoys. (Sec Reinke, in Botan. Centralblatt, 1890 ; and Oltmanns, in Pringsheim's Jahrbilcher, xxiii. 1891.) It has been commonly supposed that the composi- tion of sea-water, and particularly its degree of concentration, has a powerful influence on the dis- tribution of seaweeds. The North Sea, where the salinity reaches 3'5 per cent., is, for example, much richer in its marine flora than the Baltic — even the western part, where the salinity is T7 per cent., and still more the eastern and northern parts, where the salinity declines to 0*15 per cent. Oltmanns1 has shown, however, that the degree of salinity has much less influence than has been believed, but that rapid variations of this condition are hostile to the exist- ence of seaweeds. Where fresh water runs into the sea, it arrives in conditions varying with its abun- dance, with the currents it meets and forms, and with the direction of winds. There are thus set up differences in the density of the water, and these differences, acting on the cells of seaweeds, are of very detrimental effect. Oltmanns' observa- tions at Warnemunde, near Rostock, are of great interest in this respect. A canal there connects the sea with a lake that receives almost all the fresh- water of Mecklenburg, and many species of seaweeds grow in this lake at places where the salinity is almost nil, while almost all are absent from the canal, 1 Sitzungsber. d. K. preuss. Akad. d. Wiss. (Berlin. 1891.) 10 SEAWEEDS which conveys sometimes salt-water and sometimes fresh. The ocean currents are of primary importance as agents of distribution, not only as streams of stable temperature, but as vehicles of transport. Currents of air and of water are justly regarded as potent means of dispersal of land plants, and their efficacy in this respect is the result of special adaptation on the part of the plant, such as winged seeds, &c. No such adaptations are called for in the case of Algas towards ocean currents, though the air-floats of Fucacecc and Laminar iaccce, which secure a buoyancy in the first place for vegetative purposes, probably serve in certain cases the end of distribution as well, particularly in the gulf-weeds. As examples of the influence of currents there may be cited the differences in the marine flora between the east and west coasts of South Africa — the west under the influence of a cold stream from the south, and the east affected by the warm Mozambique current from the north ; again, the marine flora of Bermuda in the track of the Gulf Stream — the most northern coral island in the world — r-is much more markedly West Indian than the North American coast flora under the same parallel, outside the influence of this current ; and, to trace the Gulf Stream further on its path, there may be noted the contrast between the characteristically temperate marine flora of the Shetland Islands and the Arctic flora of Cape Farewell in the same parallel, but sub- ject to the cold East Greenland current. If the matter were less obvious the proof of it could be reinforced with numerous other instances. INTRODUCTION 11 The inquiry is worth prosecuting whether man affects the distribution of seaweeds. Iron vessels are much less adequately protected against fouling by the growth of seaweeds than wooden ships, which secure a considerable degree of immunity by the exfoliation of their copper sheathing, and in spite of many ingenious devices, the iron ship and steamer require frequent docking, and when sluggish, a greatly increased expenditure of coal for driving. Though cosmopolitan species like the forms of Enteromvrpha (the " grass " of seafaring language) abound near the water-line, many others, often sea- weeds of large size, occur beneath, especially when the vessel has been long at moorings. With such vessels traversing the sea in all directions, it is more than probable that the acclimatisation of aliens occurs, especially when the passage is from and to similar regions. Vessels making the passage from the Cape of Good Hope to this country across the tropics arrive with cosmopolitan forms merely, as might have been expected; but the Atlantic pas- sage between this country and North America, for example, deserves watching, and still more the Suez Canal passage. In this way man may aid the ocean currents in bridging the depths of the sea, which offer a barrier to the distribution of coast Algae. Coast deserts of sand, mud, and very friable rock in the sea are barriers frequently of great extent. There are such for ex- ample in the Gulf of Mexico from Florida to Yucatan, in the Siberian Sea, and along the muddy coast of western tropical Africa. Great irruptions of fresh- 12 SEAWEEDS water from large rivers also mark dividing lines of coast areas of distribution, and profoundly modify the character of the flora in their vicinity. The ocean forms a far less effectual barrier, however, to the dispersal of land plants than continental areas do to seaweeds, which have no means of bridging them, though many species are of sufficient range to enable them to double capes from one region to another geographically remote. The continental barrier of Africa interposed between the tropical Atlantic and the Indian Ocean offers a passage round the Cape of Good Hope, warm enough to sustain a marine flora with many subtropical types, but subject to the dis- tracting influence of opposite hot and cold currents. The result of a comparison of the floras of these two tropical regions discloses the fact, that while the genera are largely in common, the species are in a high proportion different, and this is naturally most strikingly true of orders like the Sifihonece, which are only sparingly represented outside the tropics. Areas of different temperature in the ocean have thus to be added to continental areas as natural barriers, since the ocean is never wholly fenced about with land. Such areas of different temperature, however, are effective barriers of themselves, as a comparison of the north temperate marine flora with the south temperate one shows. The heat barrier of the tropical seas would be less effective if the cold depths of the ocean were available for passage, but such depths are dark, and moreover the colder waters rise to the surface of warm seas, and thus disable the transporting action ot cold currents. INTRODUCTION 13 That there is great diversity in the marine coast floras of different regions is well known, and a comparison of three remote and dissimilar regions furnishes an extreme case. It would be difficult to select three instances of less geographical relation- ship than the Arctic Sea, the West Indies, and Australia. The first has 259 species in 111 genera, the second 788 species in 150 genera, and the third has 1,132 species in 255 genera. The Arctic Sea has 42 genera and 30 species in common with the West Indies, the same number of genera and 21 species in common with Australia, while out of the two larger totals from the West Indies and Australia there are 109 genera and 135 species in common. If we take the forms common to all three there are 32 generic types, but only 12 species out of these large totals sufficiently cosmopolitan to occur so widely. An analysis of the totals shows that in the Arctic regions the genus averages slightly more than 2 species only, while in the West Indies it is rather more than 5, and on Australian coasts rather less. The north temperate Atlantic yields an average of about 4J species to the genus, while the South African coast gives us only 3, a result which may be attributed to its small coast-line. There is still a great lack of material on which to found such comparisons with many regions of the ocean, and in the absence of full records the making of contrasts is only misleading. Enough has been said to show the diversity of such floras, the means of distribution, and the principal natural barriers that delimit the boundaries of areas. A comparison of the Arctic and the Antarctic 14 SEAWEEDS marine floras brings out the interesting result that there is a much higher proportion of forms in com- mon between the two areas than might have been expected. This is especially the case with the pelagic or free-floating plant organisms of the open sea, and without citing figures, since exact data are not avail- able, this may be taken to be generally true. Such forms are much less variable than littoral seaweeds. If we compare the two littoral floras they will be found to be of similar extent, viz. 259 species in the Arctic and 269 in the Antarctic. The Arctic species are, as has been said, in 111 genera, an average of 2 J species to the genus. The Antarctic species are in 98 genera, very nearly an average of 2f species to the genus. The genera common to both are 56, and the species 41. However, pushing the inquiry a little farther, it will be found that while some Arctic forms occur in the South Temperate zone and not in the Antarctic, similarly some Antarctic forms are found in the North Temperate zone and not in the Arctic. Adding these, we get 92 species in common to Arctic and Antarctic, including adjoining regions, and it would be much greater if we included the two Tem- perate zones fully; but the object is to compare the two cold-water floras as strictly as possible. Of these 92 species, 38 occur in the intervening tropical belt, and if they are subtracted, we get 54 species in com- mon to the two polar and adjoining waters which have not/- been found within the tropics. If we were to take the great seaweeds Fucacece and Laminariacece, the sea-wracks and tangles, we should not find even a single genus in common ; the common forms are INTRODUCTION 15 all, or nearly all, smaller seaweeds. The fact, how- ever, is sufficiently striking, that there are 54 species occurring in the two polar areas, which have been separated by a heat-belt so long as there has been climate of any sort on the globe, and if we add the even more striking resemblance of the pelagic forms, the agreement needs some theory to account for it. Marine zoologists have a similar difficulty to face. Blandet, and again recently Dr. John Murray, have brought forward the following interesting theory. In Carboniferous times, they hold, that " the surface tem- perature of the sea could not well have been less than about 70° F., and the same temperature and the same marine fauna prevailed from equator to poles, the temperature not being higher at the equa- tor. ... In early Mesozoic times cooling at the poles and differentiation into zones of climate appear to have commenced, and temperature conditions did not afterwards admit of coral reefs in the polar area, but the colder, and hence denser, water that in conse- quence descended to the great depths of the ocean carried with it a large supply of oxygen, and life in the deep sea became possible for the first time. There have been many speculations as to how a nearly uniform temperature could have been brought about in sea-water over the whole surface of the earth in early geological ages, as well as to how sufficient light could have been present at the poles to permit of the luxuriant vegetation that once flourished in these regions. The explanation that appears to me the most satisfactory is the one which attributes these conditions to the very much greater 16 SEAWEEDS size of the sun in the early stages of the earth's his- tory— an idea first introduced into geological specu- lations by Blandet (Bull. Soc. Geol. de France, ser. 2 t. 25, p. 777, 1867-68), who likewise discussed the relations of Arctic and Antarctic faunas — together with the greater amount of aqueous vapour in the atmosphere and the greater mass of the atmosphere." (Murray, in Summary of Results, " Challenger " Reports, 1895.) Another interesting point in the distribution of seaweeds bearing on this subject is that those having an incrustation of carbonate of lime occur much more plentifully in the warmer oceans — a fact equally true of the marine animals, though the pro- cess of deposition is different. This very slight development of carbonate of lime structures in the cold waters of the polar regions is instructive when compared with the massive coral reefs constructed in the polar regions in Palaeozoic and later geological times. It is a commonplace of biological knowledge that the nutrition of the animal kingdom is dependent upon the action of green vegetation in performing the primary office of converting the inorganic into the organic, and thus producing fitting substances for food. A casual observation of the great mass of terrestrial vegetation, and a comparison of it bulk for bulk with the animal life of the land, enables us to recognise the adequacy of the one as a balance to the other. On turning to the conditions that prevail in the ocean, it is at once apparent that a mere fringe of coast vegetation, extending to no great depth, can- INTRODUCTION 17 not suffice as a basis for the nutrition of the enormous mass of marine animal life which not only ranges over the great surface but penetrates into the depths of the ocean far beyond the reach of sunlight. The balance is redressed by the inconceivably great bulk of the pelagic flora, a department of the study of Algae which has been so much neglected that there is little beyond an outline knowledge of its extent. Inhabiting the surface layers of the ocean from the polar regions to the tropics there is an extensive floating marine vegetation, consisting of individuals each of microscopic dimensions, and only under special circumstances conspicuous in the mass. In the colder seas of the north and south the mass of this flora is composed of Diatomaceoc (Figs. 60), which occur in such numbers as to yield on tow- netting a palpable scum, becoming felt-like in con- sistency on drying. This living diatomaceous scum inhabits the upper layers of the waters, and rains down its dead in the form of siliceous shells on the bottom, forming extensive deposits known as diato- maceous ooze. Such deposits of marine and fresh water diatoms not only occur now on the floor of the ocean, but are preserved in rocks from the Cretaceous period, and are found, in great extent, in deposits of Tertiary and Quaternary age. While such Diatomacecc occur in greatest abundance in these regions, they have besides a wide range over the ocean surface, becoming mixed in temperate seas with Peridiniece (Fig. 56), which also are found in vast shoals. The Peridiniece are a group of organisms that require for their eluci- dation much more study than has been given to c 18 SEAWEEDS them, but there appears to be little doubt of their plant nature. Mixed with these also in temperate seas are the Coccospheres (Fig. 585), and inhabiting the warmer seas of the tropics the EhabdoepheTes (Fig. 58a), organisms of highly probable plant nature, but less studied even than the Peridiniccc. Their broken- down parts — known to geologists as Coccoliths and Rhabdoliths — are, like the remains of Diatomaccce, known from the chalk, and now play an important part in laying down the deep-sea deposits of non- polar seas, associated in this (as also in life) with the animal Foraminifera of the globigerina oozes. Min- gled with these organisms there is a profusion of pelagic Protophyta, which sometimes, as in the case of Trichodesmium erythrceum, form great banks dis- colouring the ocean over large areas, and in their origin resembling the fresh-water phenomena known as the " breaking of the meres " in Shropshire, and described by de Candolle and others as occurring in the Lake of Morat and other places, by which large sheets of water are tinged green or reddish owing to the colossal multiplication of minute fresh-water Alga?. Such occurrences have been often noted in the ocean, and, though ordinarily inconspicuous, the Algaa that cause them, and other allied forms, are always present in considerable numbers, as disclosed by the use of the tow-net. Other organisms of abundant occurrence in blue water are Pyrocystis noctiluca (Fig. 57), a source of the brilliant luminosity of tropical seas ; Halosphttra viridis (Fig. 55), of a wider range in warm and temperate seas ; and other Protococcacecv. The investigation of this pelagic flora INTRODUCTION 19 is in its infancy, and many other types no doubt await discovery. A recent estimate of the bulk of this flora compares the inconspicuous marine or- ganisms of the Sargasso Sea with the bulk of the floating banks of gulf- weed that give this great tract of ocean its name, with the result that the micro- scopic forms enormously exceed the gulf-weed in aggregate mass. The result is all the more striking since it is known that the Sargasso Sea is poor in these minute forms compared with many other regions of the ocean. As regards the Sargasso, float- ing free in this region and elsewhere, the view generally adopted accounts for their presence by the supposition that they have been torn from their natural moorings and drifted by currents, and that they slowly perish in the Sargasso Sea, to be renewed by fresh supplies from the same source. The theory has much to recommend it on purely oceanographical grounds, but the difficulty remains that Sargassum lacciferum, which composes the mass of free-float- ing Sargasso, in the tropical Atlantic, has never been recorded as growing attached, in a quantity sufficient to account for the supply. Moreover, other Sargassa do grow attached in enormous quantities, but they are of only casual occurrence in a free state. There is still the refuge that S. bacciferum is a mere "growth-form" modified by passage down currents. This, however, has no farther observation to support it, and, moreover, an examination of Sargassa from the centre of this still region of the ocean shows no symptom of recovery of broader fronds after removal from the influence of the currents that bound it. c 2 20 SEAWEEDS Though the pelagic flora is most imperfectly known as regards its constituent elements, it is manifest that its extent is enormously in excess of the coast marine flora so much more highly diversified in its forms, and that it consequently plays a role of primary importance in the economy of marine life and one of great geological interest. The distribution of Alga3 in time, as made known to us by their fossil remains, is a branch of study which is somewhat starved by the lack of material. A considerable number of so-called fossil Algae have been described by Brongniart, Saporta, and other palaaophytologists, under such names as Fucites, Chondrites, Cowfervites, Caulerpites, &c., with no better evidence of their algal nature than what may be suggested by the outlines of markings. On the other hand, Nathorst has obtained very general support for his denial of the algal nature of such markings, which he ascribes to trails of animals and other casual impressions in many cases. After weed- ing out these forms, and trusting only to such cases as exhibit microscopic structure, or characteristic casts in the round supported by evidence derived from geology as to the nature of the bed, or at the least very unequivocal impressions in beds of undoubted marine origin, there is very little left to be chronicled in the testimony of the rocks. The first appearance of Algae is in the Devonian, from which Mr. Carruthers has described Ncmatopliycus, an Alga of siphoneous structure ; and Sir Joseph Hooker Pachytheca, of more doubtful affinity as yet. With the exception of a fossil Caulcrpa from the Kimme- INTRODUCTION 21 ridge Clay, and some doubtful Dasydadacece from other beds, there is no valid evidence of other forms until the chalk is reached, with its Diatomacecc and Lithothamnion in the Senonian (Cretaceous) beds, Coccoliths, and Rhabdoliths. There come next the extensive Tertiary deposits of diatoms, and the beautiful verticillate Siphoncce of the same age described by Munier Chalmas ; the coralline Litho- thamnion and the fresh-water Characecc. The later beds furnish little of interest except Quaternary diatornaceous remains. There is room for dis- appointment in the failure to find indubitable records in earlier rocks of plants of such primitive type as the Algae, and it is startling to find such forms as the diatoms suddenly burst upon geological history in a profusion of genera and species, many: of which survive in their specific forms from their first appearance to this day. From all that is known of the Silurian rocks, for example, the discovery of diatoms in them would appear to be highly probable, but research has failed to discover them. However intractable and therefore suited to preservation their siliceous shells may be, the existence of conditions under which this substance would become fugitive is probable, and the gap, though significant enough, is not more so than the absence of Muscincce, for example, from the coal measures. It but emphasizes the imperfection of the geological record of plant history, and points to caution in generalising from an insufficient array of facts, more than it indicates argument in favour of any particular sequence of primitive plant forms. 22 SEAWEEDS The conditions of environment of seaweeds are, as has been described, by no means so complex as those of land plants, and their general adapta- tion to their surroundings is expressed in a cor- responding simplicity of structure. The aquatic habit, fresh -water and marine, is accompanied in flowering plants by a degradation of structure in their vegetative organs, since the buoyancy of water, aided by the air-spaces of the plants, dispenses with the need of the mechanical aid of vascular tissue, and partly of its conducting function. This tissue is accordingly much reduced in aquatic flowering plants, and there is a corre- sponding reduction in the epidermal system, since there is no need of a special cuticular or corky layer to protect the plant from undue evaporation. A favourite view of the evolution of plant forms represents their ascent as a process of gradual emancipation from an aquatic habit; and the adoption of this habit by members of highly de- veloped groups as of the nature of a relapse or approximation to their primitive state. The student of seaweeds is not concerned with the point farther than it is founded on the fact that environ- ment has made no demand on these organisms of a kind that calls for much specialisation of their tissues to enable them to adapt themselves to it, and throughout the group there is a simplicity of structure and a plasticity of form of express character. In the most highly organised seaweeds the vegetative tissues may be classified into a cortical assimilative layer and a central conducting INTRODUCTION 23 strand, the mostly highly developed element of the latter being the sieve-tubes found in Macrocystis (Laminariacecv). The sculpturing of outward form reaches its highest point in the differentiation (1) of a root-like holdfast, which, however, is not an organ of absorption unless possibly in the case of certain partially parasitic forms (cf. Notheia, Choreocolax, Figs. 5rf, 64) ; (2) of a stem and (3) of leaf-like appen- dages. From this type there are varying intermediate forms down to the wholly undifferentiated type, which occurs among both multicellular and unicellular forms. These intermediate forms may be placed into two categories — those exhibiting a root-like differentiation from an otherwise unspecialised body, and those in which there is merely a distinction between base and apex. In some of the lower multicellular Algae all the cells are alike, and equally capable of vegetative and reproductive functions. Among the unicellular forms there are those which exist free singly, and others united into a kind of spurious tissue or colony by a common investing mucilaginous cell-wall, and occurring either in rows or filaments, or in more or less indefinite masses. The highest development attained by the unicellular forms, if they may be so termed in this connection, is to be found in the multinucleate group of Siphonece, which includes many forms with differen- tiated root-like appendages and leaf-like shoots, and others in which the specialisation is carried so far as to represent leaf-like, stem-like, and root-like organs. Growth in length is either (1) apical, and effected 24 SEAWEEDS by a single apical cell, or a marginal series, or a meristematic group in the multicellular forms, or by the apical protoplasmic contents in siphoneous plants ; or (2) intercalary, at a definite growing-point in the frond, as in Laminariaceaf, &c., or in a terminal hair or tuft of hairs with a basal growing- point. Sometimes all the cells of the body remain meristematic and engaged in growth. Secondary growth in thickness may take place, as in the stalk of Laminaria, by the peripheral cells beneath the rind being capable of division, and thus adding to the internal tissues, as well as forming towards the outside a bark-like rind ; or by an adventitious process, as in Desmarestia, where the branching filaments grow together into a kind of pseudoparen- chymatous tissue, and invest the original cellular axis. The cohesion of the body is effected in various ways; either by the union of the cells into a parenchymatous tissue, or by the intertwining of filaments, aided in some cases by the development of sucker-like holdfasts called tenacula or haptera (Udotea, Struvea, &c.), or by incrustations of carbon- ate of lime (Corallinece, Squamariecc, and Siphonece). Stability is obtained in Caulerpa by the formation of numerous trabeculse or branching cellulose cross- beams, braces, or struts, traversing the interior of the cell-cavity from one part of the wall to another, and enabling this remarkable Alga to assume a differentiation into leaf-like, stem-like and root-like parts, though it consists of but one great cell- cavity. Intercellular spaces are most prominently INTRODUCTION 25 represented by the air-bladders which secure buoyancy for the Fucaccce and Laminariacece, while the whole interior of many other forms is hollow. The mucilaginous character of the cortical tissue of many Algae protects the internal cells from drying while exposed between tides. In the rind of the Laininariacece there are special gum -passages, while in Splachnidium and other forms the whole of the interior is filled with a mucilaginous substance. Adhesion to the substratum is effected by sucker- like haptera, by basal layers of cells, or by rhizoid filaments which penetrate the substratum. A comparative review of the reproductive pro- cesses of seaweeds would be unprofitable by itself, since such a treatment would lack symmetry without reckoning in the fresh-water forms. It would be, moreover, appropriate only in a treatise on the general morphology of all Algae. Details of such processes are given in the accounts of the differ- ent groups, but it is of interest to note now the occurrence in seaweeds of isogamous and ooga- mous forms of reproduction, and the propagation by spores and other bodies of purely vegetative character. Though these modes of reproduction are represented in their typical forms among seaweeds, certain subordinate types are confined to the fresh- water Algae. Conjugation by non-ciliated gametes, for example, occurs in the sea only among the Diatomacecc, since the Gonj-ugatce, of which it is characteristic, are all fresh- water Algae. There appears to be an almost equal amount of diversity of type of reproduction among fresh-water and marine 26 SEAWEEDS forms — if anything, fresh-waters are richer in such types, while on the other hand the sea is incompar- ably more favourable to diversity of vegetative development and to luxuriance of habit as well. It was at one time supposed that among the red and green Algae there were forms in which the chlorophyll was diffused throughout the protoplasmic cell-contents, but research has shown that in all cases examined in these groups and in the Phwophycece as well, true chromatophores occur, while they are absent, so far as is known, from the Cyanophycecc. These chromatophores (chlorophores, erythrophores, phoeophores or melanophores, as the case may be) sometimes contain pyrenoids — minute bodies (appearing within them much as a nucleolus appears within a nucleus) confined to the chromato- phores of Algae with the single known exception of Anthoceros (Hepaticae). The chromatophores occur either singly in each cell or in numbers, and are of definite and characteristic shapes. These shapes are not only of constant character, but the same constancy extends to the fact of their single or numerous occurrence in each cell. The presence or absence of pyrenoids, which may vary in size from time to time, affords a more capricious character, since forms possessing them are found among the PhceopJiycece, Rhodophycece, and Chlm^opliycew. Various attempts have been made to attach special significance to the occurrence of pyrenoids, but so far there has not been much success in elucidating this point. There is, for example, no clear ground for the view that their presence is connected with INTRODUCTION 27 an incomplete differentiation of reproductive and vegetative cells in the plants which possess them. It may be so, but the point is by no means established. The special amyloplastic function of the chromatophore, as distinguished from its assimilative one, appears to be limited to that portion immediately investing the pyrenoid. If the change of size of the pyrenoids be associated with the nutritive state of the cell, as appears probable, it would con- firm the opinion that they are reserves of proteid. In collecting seaweeds between tide-marks the nature of the appropriate equipment is so obvious as to need little direction. The most convenient receptacle for specimens is the ordinary waterproof sponge-bag, though a tinned iron milk-can with a good lid has its advantages. In no case should glass bottles be carried in the hand or pocket, since they are a source of danger, and unnecessary. For similar reasons a knife is an undesirable companion in slippery places, and it is not needed if the collector carry a stout stick of the alpenstock pattern, with a chisel screwed into one end for scraping off specimens, and a small landing-net ring provided with a cotton bag instead of a net attached to the other end. The stick is useful for support, and the bag for securing floating specimens that have been detached by the chisel. Wading boots are of great advantage, except where there are deep pools, when the risk attending immersion is greatly increased by their use. Though a storm is often more productive than a dredging expedition, the specimens are frequently SEAWEEDS much lacerated, and should be collected at once to be of any value, since exposure quickly spoils them. The ordinary form of dredge used in securing zoological specimens serves for seaweeds, but is liable to become choked. Reinke's dredge, armed with cutting teeth like spear- heads surrounding the mouth, has been found to be service- able, and a simple contrivance shaped like the letter A, with strong fishhooks of the largest size on strong cords attached to the cross-bar, and the ends of the legs weighted, has been recom- mended, but requires skill and judgment in its use. It is likely to disappoint the experimenter unless under very favourable con- ditions. A light dredge can be easily worked from a rowing- boat, which is also sufficient for no. i.-Reinke-s Dredge, tow-netting with fine silk nets for the capture of the free-float- The traveller who wishes to examine the minute pelagic Algse from the surface layers of the ocean can do so by obtaining permission to tap an inlet pipe of a steamship, and allowing the water to run through a fine silk bag for a time, when he will be rewarded with results similar to those from tow-netting. This method has been success- fully practised by its inventor, Dr. John Murray, of the Challenger expedition. ing minute Algse. INTRODUCTION 29 In drying specimens the material should be floated out, and a mount-paper of suitable size placed under it and slowly lifted out by one corner. By means of a camel-hair brush the branches may be kept apart, since they are apt to become entangled at the critical moment of leaving the water. When this happens at one or two spots merely, a drop of water placed on the part will permit of rearrangement without plunging the whole in again. A number of specimens may be dried simultaneously by using, instead of a basin, a shallow zinc tray with a per- forated or wire-woven plate large enough for several specimens. It requires practice in lifting it out, and though specimens good enough for botanical pur- poses may be so obtained, they are never so beauti- fully arranged as when taken out singly on their mounts. The wet specimens on their mounts should be placed at once between sheets of drying- paper (blotting-paper is too absorbent) with a layer of muslin over each sheet of specimens to prevent their adhering to the upper sheet of drying-paper. As a rule seaweeds need less pressure than flower- ing plants, and the collector will very soon learn to adjust it. Plenty of drying-paper should be used, and frequently changed — twice during the first twenty-four hours, and once afterwards until the specimens are quite dry. Though dried specimens can be easily soaked out again for microscopic examination, they are never so good for this purpose as those that have been preserved in fluid. A good method of soaking is to place the part to be examined, over-night or longer, in absolute 30 SEAWEEDS alcohol, to remove as much air as possible. It should then be transferred to salt and water, and permitted to remain in it. A drop or two of glycerine should be added, and the process may be hastened by gently heating, not boiling. The most successful specimens are those that have been kept at about 90° F. for several hours. Material so treated may then be preserved in spirit (at first weak, and gradually strengthened). Living specimens to be preserved in spirit should be first treated with picric acid. A saturated solu- tion of picric acid in sea-water should be made and subsequently diluted with three or four times its volume of sea-water. The specimens should be im- mersed in it from a quarter of an hour to two hours, according to size and density ; then washed in sea- water and placed first into weak spirit, and by degrees into stronger spirit. Specimens vary greatly as to the result of treatment by picric acid. To ob- tain thorough fixation of the contents of such Algae as Valonia (Fig. 46) it is necessary to immerse them for several hours in the saturated solution itself. For mounting microscope slides of seaweeds the best medium is clear glycerine jelly, which has the advantage of being easily manipulated. The examin- ation of Algae encrusted with carbonate of lime, such as the stony corallines, is facilitated by the use of Perenyi's decalcifying fluid (4 vols. 10 per cent, nitric acid, 3 vols. absolute alcohol, and 3 vols. 5 per cent, chromic acid), which gives better results than weak hydrochloric acid or any other method in common use. It is particularly valuable in examining Algae INTRODUCTION 31 like Gomontia (Fig. 50), which bore into and inhabit shells. The whole of the investing substance may be thus removed, and the Alga disclosed without any breaking up of its filaments or injury to its cells. In other respects the ordinary methods of microscopical examination are sufficient. In considering the economic uses of seaweeds, the indirect service they render as the basis of the nutri- tion of animal life in the sea, and consequently their fundamental importance for fishery, must not be left out of account here, as it practically has been by fishery boards and others whose main concern it might appear to be. In investigating the food of fishes, the so-called practical inquirer is accustomed to look no further than the immediate organisms eaten, much as if in agricultural matters no heed were given to the pasturage of farm stock. There is no doubt that the enormous shoals of Peridiniece and other allied free-floating Alga3 are the pastures on which the organisms constituting the food of fishes themselves feed — that in fishery matters they are the basis of the pyramid of which man is the apex, and the dearth of knowledge of these forms and the indif- ference of fishery authorities to the subject in its technical aspect, is only equalled by the ignorance and apathy of botanists towards its scientific value. The direct economic importance of Algae is no longer so great as it was early in the present cen- tury, when the kelp industry flourished in the north of Scotland and the western coasts of Scotland and Ireland. The value of kelp in the manufacture of soap and glass became greatly enhanced by the 32 SEAWEEDS exclusion of barilla as an import from our markets during the long war. The price of kelp then rose so high that the income of the Outer Hebrides from this source alone was computed to have reached £120,000 a year. The industry practically came to an end with the peace and the reintroduction of barilla, while the removal of the salt duty struck a further blow at the revenue of these districts. Kelp is now used only in the manufacture of iodine, and as a manure. The common sea-wrack or bladder- wrack (Fuciis vesiculosus) has been used medicinally for a variety of diseases, but its reputation in this respect has been acquired principally as a remedy for obesity. Bentley and Trimen (Medicinal Plants, vol. iv. p. 304) state that " farther trials are neces- sary before any definite conclusions can be arrived at on its action, and its value as a remedy in obesity. It would appear that it is the essential constituent in the nostrum now so extensively advertised under the name of Anti-Fat." As a food, or rather as a sauce, the species of Porphyra known as Laver are not sufficiently appreciated. Laver is not only abun- dant, but is easily preserved. Carragheen, sometimes called Irish Moss (Chondrus crispus), is used for its nutritive properties, which however appear to have been over- valued. Dulse (Rhodymenia palmate) and tangles (Alaria escidentd) are eaten by the hardy, but are extremely indigestible. The Chinese and Japanese engage in an extensive industry in seaweed products, and certain species are cultivated. Ceylon Moss, or Jaffna Moss (Gracilaria lichencddes), a seaweed which abounds in Eastern seas, is the source of Agar-Agar, INTRODUCTION 33 so useful in the preparation of culture-media for observing the growth of bacteria. The same species, together with another (Gracilaria confervoides, and probably also species of Eucheuma) are used by the Chinese and Japanese for making jellies and sweet- meats, and for stiffening purposes, varnishes, &c. The extensive deposits of fossil Diatoms which occur in several quarters of the world furnish in some cases valuable polishing powders, and are used, as in the case of the Hanoverian Kieselguhr deposit, for mixture with nitroglycerine to form dynamite. 34 SEAWEEDS The following selected books and papers will be useful for reference, and will guide the student to farther literature. SYSTEMATIC. General. AGARDH, J. G. Species, Geuera et Ordines Algarum. 3 vols. 1848-80. Leipzig. AGARDH, J. G. Till Algernes Systematik. 1872-99. Lund. KiiTziNG, F. T. Tabulae Phycologicee. 19 vols. 1845-69. Nord- hausen. DE TONI, J. B. Sylloge Algarum. Padua. (In progress. ) GEOGRAPHICAL. Britain. GREVILLE, R. K. Algae Britannicae. 1830. Edinburgh. HARVEY, W. H. Phycologia Britannica. 4 vols. 1846-51. London. BATTERS, E. A. L. Marine Algce of Berwick-on-Tweed. 1889. Alnwick. HOLMES, E. M. , and BATTERS, E. A. L. A Revised List of the British Marine Algae. 1892. Oxford Univ. Press. Europe. HAUCK, F. Die Meeresalgen Deutschlands und Oesterreichs (in Rabenhorst's Kryptogamen Flora). 1885. Leipzig. REINKE, J. Atlas deutscher Meeresalgen. 1889-92. Berlin. ZANARDINI, G. Iconographia Phycologica Adriatica. 1860-76. Venice. N. America. HARVEY, W. H. Nereis Boreali-Americana. 3 parts. 1851-58. Washington. (Including Atlantic and Pacific coasts.) FARLOW, W. G. Marine Algae of New England. 1881. Wash- ington. Arctic Sea. KJELLMAN, F. R. The Algae of the Arctic Sea. 1883. Stock- holm. (See also the author's Beringhafvets Algflora. 1889. Stockholm. ) West Indies. MURRAY, G. Catalogue of the Marine Algae of the West Indian Region. 1889. London. INTRODUCTION 35 Cape of Good Hope. HARVEY, W. H. Nereis Australis. London. 1847. (Refers to other Southern Ocean regions besides the Cape of Good Hope. ) BARTON, E. S. A Provisional List of the Marine Algae of the Cape of Good Hope (Journal oj Botany, 1893). Fuegia. HARIOT, P. Algae in Mission Scientijique du Cap Horn. 1889. Paris. Australia. HARVEY, W. H. Phycologia Australica. 5 vcls. 1858-63. London. SONDER, W. Die Algen des tropischen Austr aliens. 1871. Ham- burg. BONDER, W. List in Baron von Mueller's Fragmenta Phyto graphics Australia?, vol. xi., Supplement, and also in Addi- tamenta. Indian Ocean. HARVEY, W. H. List of Ceylon Algae, with distributed set oi specimens. Tropical Pacific. HARVEY, W. H. List of Friendly Islands Algae, with distributed set of specimens. North Pacific. RUPRECHT, F. J. Neue oder unvollstandig bekannte Pflanzen aus dem nordlichen Theile des Stillen Oceans. 1852. St. Petersburg. POSTELS, A., and RUPRECHT, F. Illustrationes Algarum. 1840. St. Petersburg. SURINGAR, W. F. R. Algae Japonicae. 1870. Haarlem. On the general subject of distribution see also— MURRAY, G. The Distribution of Marine Algae in Space and in Time (Trans. Biol. Soc. Liverpool, vol. v. ). MURRAY, G. A Comparison of the Marine Floras of the Warm Atlantic, Indian Ocean, and Cape of Good Hope (Phy co- logical Memoirs, Part II. ). MURRAY, G., and BARTON, E. S. A Comparison of the Arctic and Antarctic Marine Floras (Ibid. Part III. ). D 2 36 SEAWEEDS MORPHOLOGY. General. THDRET, G., and BORNET, ED. Etudes Phycologiques. 1878. Paris. BORNET, ED., and THURET, G. Notes Algologiques. 1876-80. Paris. DERBES, A., and SOLIER, J. J. Mem. de la Physiologic des Algues. 1856. Paris. WILLE, N., and KJELLMAN, F. R. Algae in Engler's Naturlichen Pftanzenfamilien. (In progress.) FALKENBERG. Die Algen, in Schenk's Handbuch der Botanik, vol. 2. 1882. ARESCHOUG, J. E. Observationes Phycologicre. Upsala. 1866-75. MURRAY, G. Phycological Memoirs. 1892-95. London: Dulau and Co. SCHUTT, F. Das Pflanzenleben der Hochsee. 1893. Leipzig. SCHMITZ, F. Die Chromatophoren der Algen. 1882. Bonn. Phceophycece. BOWER, F. 0. On the Development of the Conceptacle in the FucacecK (Quart. Journ. Micr. £a.,1880). OLTMANNS, F. Beitrage zur Kenntniss der Fucaceen. 1889. Cassel. BERTHOLD, G. Die geschlechtliche Fortpflanzung der eigen- tlichen Phaeosporeen (Zool. Stat. Naples, vol. 2, 1881). JANCZEWSKI, E. Observations sur 1'accroissement du thalle des Phaeosporees (Mem. Soc. Nat. d. Sc. Cherbourg, 1875). SETCHELL, W. A. On the Classification and Geographical Distri- bution of the Laminariaceae (Trans. Connecticut Acad., 1893). KARSAKOFF, N. Quelques Remarques sur le genus Myriotrichia (Journ. de Bofanique, December, 1892). VALIANTE, R. Le Cystoseirag del Golfo di Napcli (Zool. Stat. Naples, 1883). BARTON, E. S. A Systematic and Structural Account of the Genus Turbinaria ( Trans. Linn. Soc. Bot. , new series, vol. 3, 1891). Chlorophyceie. SCHMITZ, F. Halosphcera, eine neue Gattung, &c. (Mittheil. Zool. Stat. Naples, 1879). ARESOHOUG, J. E. Letterstedtia, ny Alg-form fran Port Natal (Ofvers af. Vet. Akad. Fo'rhandl. Stockholm, 1850). BORNET, ED., and FLAHAULT, CH. Sur quelques plantes vivant dans le test calcaire des Mollusques (Bull. Soc. Bot. France, xxxvi. ). WORONIN, M. Beitrage zur Kenntniss der Vaucherien (Bot. Zeit. INTRODUCTION 37 SOLMS-LAUBACH, GRAF zu. Monograph of the Acetabulariese (Trans. Linn. Soc. Bot. 1895). SOLMS-LAUBACH, GRAF zu. Cymopolia, Neomeris, and Bornetella (Ann. Jard. Sot. Buitenzorg, 1892). CRAMER, C. Neomeris and Cymopolia (Denkschr. d. Schweiz. Naturf. Gesellsch. xxx.). CRAMER, C. Neomeris and Bornetella. (Ibid, xxxii. ). MURRAY, G. On Boodlea (Journ. Linn. Soc. Bot. vol. xxv.). MURRAY, G. , and BOODLE, L. Struvea (Annals of Botany, vol. ii. ). MURRAY, G., and BOODLE, L. Avrainvillea (Journal of Botany, 1889). CORRENS. C. Ueber die Membran von Caulerpa (Ber. Deutsch. lot. Gesellsch. vol. xiiA WEBER-VAN BOSSE, A. Etudes sur des Algues de PArchipel Malaisien (Ann. Jard. Bot. Buitenzorg, 1890). Fossil Chlorophycew. SOLMS-LAUBACH, GRAF zu. Einleitung in die Palaeophytologie. 1887. Leipzig. For literature of Diatomacege, see De Toni, Sylloge, vol. ii. See also Pfitzer in Schenk's Handbuch der Botanik, 1882, for structure, &c. Rhodophycece. SCHMTTZ, F. Untersuchungen iiber die Befruchtung der Flori- deen (Sitzber. K. Acad. Berlin, 1883). SCHMITZ, F. Floridese in Engler's Syllabus der Vorlesungen ilber Botanik, 1892. Berlin. SCHMITZ, F. Systematische Ubersicht der bisher bekannten Gattungen der Florideen (Flora, 1889). SCHMITZ, F. Die Gattung Actinococcus (Flora, 1893). SCHMITZ, F. Kleinere Beitrage zur Kenntniss der Florideen (La Nuova Notarisia, 1892-94). HAUPTFLEISCH, P. Die Fruchtentwickelung der Gattungen Chy- locladia, Champia, und Lomentaria (Flora, 1892). DAVIS, B. M. Development of the Frond of Champia parvula from the Carpospore (Annals of Botany ', vol. vi.). HEYDRICH, F. Pleurostichidiitm, ein neues genus der Rhodo- meleen (Ber deutsch. Bot. gesellsch. vol. xi. 1893). ZERLANG, 0. E. Die Florideengattungen Wrangelia und Nac- caria (Flora, 1889). JOHNSON, T. Stenogramme interrupta (Annals of Botany, vol. vi. ). RICHARDS, H. M. On the Structure and Development of Choreo- colax PolysiphonivK Reinsch (Proc. Amer. Acad., vol. xxvi. ). KUCKUCK, P. Choreocolax albus, ein echter Schmarotzer unter den Florideen (Sitzber. K. preuss. Akad. Wiss. Berl. xxxviii. 1894). 38 SEAWEEDS BERTHOLD, G. Die Bangiaceen des Golfes von Neapel (ZooL Stat. Naples, 1882). Cyanophycece. BORNET, ED., and FLAHAULT, CH. Revision des Nostocac^es he'teYocyste'es (Ann. Sci. Nat. Bot. ser. vii. vols. 3, 4, 5, and 7). GOMONT, M. Monographic des Oscillariees (Nostocacees homo- cystees) (Ibid. vols. 15 and 16). ZACH ARIAS, E. Ueber die Zellen der GyanopJiyceen (Bot. Zeit. 1892 and 1893). Murray's Seaweeds Plate I Han hart imp PLATE I. 1. PELVETIA CANALICULATA. 2. HALIDUYS SILIQUOSA, 3. CYSTOSEIRA ERICOIDES. 4. CtTTLERIA MULTIFIDA. SUB-CLASS I PH^OPHYCE^E WITH the exception of some of the species of Lithoderma and the genus Pleurocladia, represented only by a minute form of doubtful affinity dis- covered by Alexander Braun in the Tegeler See near Berlin, all the Phceophycece, or Melanophycece as they are otherwise called, are seaweeds. They agree in the fact that all their motile reproductive cells, zoospores, antherozoids, and gametes are pro- vided with two lateral cilia, one pointed forwards and the other backwards in motion ; in the fertilisa- tion of their oospheres and the conjugation of gametes outside the parent plant, and the direct germination of the zygote which is the product of this union ; in the possession of brown chromatophores tinged with phycophaeine and phycoxanthine (the phyco- phaeine being soluble in water and the phyco- xanthine in alcohol, the compound pigment being termed phaeophyll); and in having mostly but one nucleus in the vegetative cells. The vegetative body of the plants coming under this sub-class is of great diversity, including the most highly organised of all seaweeds, of giant dimensions and 40 SEAWEEDS great external differentiation of form and consider- able internal differentiation of tissue, as well as others consisting of a mere row or plate of similar cells. The sub-class may be regarded as a fairly natural assemblage of orders easily to be distin- guished from the other sub-classes, though it includes such diverse types as (1) the Fucacece, of which the unciliated oospheres, many thousand times greater than the antherozoids, are produced like the latter in definite conceptacles, from which they are ex- truded; (2) the Cutleriacece, possessing non-sexual zoospores, and ciliated oospheres (or $ gametes) many times larger than the antherozoids (or -i. — Kjellmania sori- fera. Filament with young sorns sporangia. Highly mag- nified. (After Rcinke ) FIG. 26. — Stictyosiplion DecaisneiL Transverse section of thallus showing sporangia. Highly magnified. up by walls running transversely to the axis of the plant into a sorus of cells, each of which ac- quires an arched apex and divides farther into 2-4 loculi one above the other. Each loculus produces a single zoospore. The whole sorus is thus in this case the product of a single superficial cell. The occurrence here and in Giraudia of two kinds of 104 SEAWEEDS plurilocular sporangia leads to a certain amount of hesitation in adopting the presumption that these always contain gametes or even potential gametes. It recalls the difficulty presented by the case of Ectocaiyms secundus, which possesses plurilocular sporangia and antheridia (sec p. 68). Zostcrocarpus, a genus recently founded by M. Bornet, appears to be most nearly related to Kjellmania, especially in the mode of formation of the sori of sporangia. The Geographical Distribution is mainly in the North Atlantic and Arctic oceans. Stictyosiphon occurs in the Mediterranean and South Atlantic, Kjellmania in the Baltic, Striaria in the North Atlantic and Mediterranean, and Ptdctospora in the Arctic and North Atlantic. Striaria and Phl-ceospora are British. Some authors regard species of Pklceospora, probably correctly, as belonging to Sticty- osiphon (e.g. Phlceospora tortilis), which accounts for the presence of that name in British marine floras, while others restrict Stictyosiphon to the single species S, adriaticus. ENCCELIACE.E. General Characters. — Though none of the Enccdi- acece attain great size, the order is remarkable for the great diversity of the forms assumed by the thallus, including frond-like, filamentous, club-shaped, glob- ular, hollow, reticulate, &c. shapes, though they agree in none of them having a definite system of branching. They arc of parenchymatous structure and PH^OPHYCE^E 105 possess no definite growing point. The reproductive organs are both unilocular and plurilocular sporangia arising by the differentiation of a superficial cell or of an outgrowth from one. The conjugation of gametes has been observed in one form. The Thallus in the filamentous form is articulated, and consists of several rows of cells, but sometimes partially or even wholly of a single row. However, most of the forms exhibit both a cortical tissue and an internal one. The internal tissue consists of large parenchymatous cells, and the cortical layer of smaller assimilative cells, though these are commonly of relatively greater size than in allied groups. Hairs and paraphyses sometimes spring from the super- ficial cells, the hairs with basal growth (at least in the case of Punctaria, and more notably Hydrocla- thrns) from pits resembling cryptostomata of simple structure. The attachment of the thallus to the substratum is either by means of a disc or by a weft of rhizoids. Immediately above the attachment the base of the thallus is commonly attenuated to a thin, solid and short stalk. There is no true growing point, and the growth is distributed more or less equally over the whole thallus, but persists at the base, as a rule, for a time after it has ceased elsewhere. The development of the cryptostomata has been observed in Hydrodathrus. In a surface view " an isolated cell or several cells in a group become separated off from the surrounding epidermis, each loses its polygonal shape and becomes cylindrical. ... In a radial sec- tion of such a group each cell is seen to be divided by a transverse wall, but there is no indication of 10G SEAWEEDS such longitudinal division as occurs in neighbouring epidermal cells. The lower of the two cells again ft FIG. 27. — a, Hydraclathru* cancellatux one-half natural size ; b, section of thallus ; c, H. sinnomts, cryptostoma and paraphyses : d, mature cryptostoma with young sporangia, b, c, d highly magnified. (After M. O. Mitchell, in Phyc. Mem.) divides transversely, and this method of division continues till a long row of cells has been formed, making in fact a hair. Simultaneously with the 107 formation of these hairs, the cells immediately sur- rounding them undergo similar changes, and thus the cryptostoma enlarges radially. Meanwhile the thallus continues its growth, so that the basal cells of the hairs which were originally in the same plane as the epidermis have now come to lie below it, and the whole structure is suggestive of a conceptacle." L It is interesting to note that the formation of a crypto- stoma is the starting point for the formation of a sorus of plurilocular sporangia from the adjacent epidermal cells. This spreads radially from the cryptostoma as from a centre, and the formation of sporangia is suc- ceeded, after these have disappeared, by that of club- shaped paraphyses from the basal cells that bore the sporangia. While the basal cells nearest the crypto- stoma are producing paraphyses, those farthest away are still giving rise to sporangia. Gradually the paraphyses replace the sporangia until the latter disappear, and there is left a group of paraphyses with a central cryptostoma. This occurrence of a central cryptostoma in the sorus recalls the case of Adcnocystis2 (Laminaricicccv, p. 84), though it presents a contrast with the conceptacle of Splachni- dium, which bears hairs only at first, and sporangia later on. Though not comparable with the more highly developed cryptostomata of the Fucacccc, we have here an elementary form of cryptostoma, and it is instructive to observe that the development of cryptostoma and sorus originates in the alteration of 1 M. O. Mitchell, in Murray's Phycoloyical Memoirs, part ii., p. 54. 2 Murray in Fhyc. Mem., part ii., p. 62. 108 SEAWEEDS form and division of one initial cell (or at most a small group) derived from the epidermal layer. The Reproductive Organs, both unilocular and pluri- locular, are either differentiated superficial or epi- dermal cells or outgrowths of these. They are accompanied or succeeded, as has been noted above, by the formation of paraphyses. While in such genera as Desmotrichum, Punctaria, &c., the epi- dermal cells in question undergo but little differ- entiation, in Scytosiphon, Hydrodathrus, and other genera, the cells giving rise to plurilocular sporangia undergo considerable elongation and division in the process. The globular unilocular sporangia of Asper- ococcus, standing free from the surface of the plant among paraphyses, originate in outgrowths from the epidermal layer — as the paraphyses do. The order is formed of three others recognised formerly as Punctariaccm, Scytosiphonacew, and Asper- ococcaccce, together with the genera Ooilodesme and Myd&phycus — a grouping of them justified by Kjell- man (Engler and Prantl's Nat. Pflanzcnfamilicn, part i., p. 197). The Geographical Distribution is fairly even through- out all seas, but more abundant in temperate waters. Desmotrichum, Punctaria, Litosiphon, Scytosiphon, Phyllitis, and Asperococcus occur in British seas. RALFSIACE^:. General Characters. — This small order embraces plants of a very undifferentiated vegetative structure and of a crust-like habit. The reproductive organs 109 are of two kinds, plurilocular (containing pre- sumptive gametes) and unil ocular. The Thallus of Ralfsia forms leathery crusts on rocks, &c., and though at first almost circular, be- comes ultimately of irregular outline. It attains considerable thickness in the more central portion, and decreases towards the margin. The cells, each with one chromatophore,are in vertical series and form a parenchymatous tissue ; those of the margin effect- ing by division the extension of the thallus, while the superficial cells similarly add to its thickness. It is attached to the substratum either directly or by means of root-hairs. On the upper surface there occur single hairs, or in other cases tufts, sometimes springing from pits which, however, do not appear to exhibit a special development like the cryptostornata of other orders. By a process of overlapping of new crusts on older ones, the thickness of the thallus is often considerably increased. In Lithoderma there is no notable difference from Rcdfsia in the develop- ment and structure of the thallus. The Reproductive Organs. — Until recently only unilocular sporangia were known in the case of Rcdfsia. They are obovate and ' arise as lateral processes from septate hairs, which are in turn direct prolongations of the superficial cells of the thallus. Though of lateral origin, the sporangia assume a terminal position by pushing aside the true terminal shoot. The plurilocular sporangia, recently discovered by Mr. Batters, are of similar origin, but with them are no paraphyses. The hairs or paraphyses grow mainly, if not wholly, 110 SEAWEEDS by the division of their apical cells, and, at the margin of a sorus, where they may be seen in an early stage of development, push off and cause to exfoliate the original cuticuloid gelatinous layer of the epidermis. In Lithoderma sporangia of both kinds occur in sori, but on different plants. The plurilo- cular bodies, which may be presumed to give rise to gametes, occur as lateral off-shoots from special branches arising in turn from superficial cells. These gametangia are either a single row of cells, or more frequently several rows, and are more or less cylindrical in shape. The unilocular sporangia are terminal bodies — in fact transformed superficial cells, and are mostly obovate. The actual life-history has not been followed, but the relationship with Ealfsia cannot be doubted. The general relation- ships of the order are vague, but probably Chor- dariacece of which Ralfsiacece may be degenerate allies, and Ectocarpacem (especially Ascocychui) exhibit the most distinct claims. The Geographical Distribution extends from the polar seas to the tropics, and though the species are always few, the maximum is attained in the North Atlantic. The distribution of the six or seven species of Lithoderma is of limited range so far as is at present known. They occur, however, both in fresh-water and in the sea. The marine forms are found in the Arctic Sea, North Atlantic, and Mediterranean, while the two fresh- water forms occur in the south of France, Germany, and Sweden. L. fatisccns has recently been discovered on the British coasts. PH^EOPHYCE.E 111 SPHACELARIACE.E. General Characters. — The thallus consists of erect shoots springing from a more or less extensive basal creeping cushion fixed to the substratum. The erect shoots are more or less branched and may be as in the most reduced case a simple cell-row, or an articu- lated filament consisting below of tiers of cells of equal height, or the latter case may be further developed by the addition of a cortical tissue. Growth in length is effected by an apical cell. The reproductive organs are both unilocular and pluri- locular and are borne on short branches. Vegetative propagation by gemmae occurs. The Thallus. — The most simple form is to be found in the exceptional case of Battersia mirabilis, for the inclusion of which the definition of the order has to be somewhat stretched, since this plant exhibits in many respects a striking resemblance to Litho- derma, and in these same points a divergence from Sphacelariaccw. The thallus of Battersia consists of a creeping cushion of several layers of cells fixed to the substratum by its undersurface,and giving off upwards short simple or branching shoots which bear the sporangia terminally. These shoots consist of single cell-rows, though occasionally they form several rows at the base, after longitudinal division of the cells. Most of the SphacclariacecK spring from a basal cushion which grows by its marginal cells, or from a weft of rhizoids which penetrate the tissue of the 112 SEAWEEDS host on which they may be epiphytic. Runners proceed from these, which sometimes give rise to new plants. The erect shoots are mostly articulated - I II tig U II filaments consisting of tiers of cells and terminating upwards in an apical cell, by division of which growth takes place. This apical cell is of relatively large size and the shoot possesses the remarkable power PH^EOPHYCE^E 113 of being able to renew the apical cell after the original one has perished. This property is akin to another exhibited by the branches which bear the gemmae and sporangia. After the fall of a gemma, the remaining basal cells of the branch that bore it may proceed to grow and form another one. Similarly after the emptying of a sporangium, the branch may grow through the membrane and form a new sporangium. The branches are either all of equal morphological value or there may be a distinction between long and short ones. The latter are alternate in Stypocaulon and Halopteris, opposite in Chcetopteris, and in whorls in Cladostephus. They are either themselves unbranched (Stypocaulon and Chcetopteris), or branched on one side (Cladostephus), or they re- semble in miniature the branching of the long branches (Halopteris). In Stypocaulony Cladostephus, and not so markedly in Halopteris, the outermost cells of the filaments of the thallus undergo farther division and produce a cortical tissue of small cells which obscures the articulation of the original thallus. In Sphacelaria and Chcetopteris this articulation of tiers of cells remains apparent throughout the life of the plants. A farther modification takes place in Cladostephus, Stypocaulon, Halopteris, and some species of Sphacelaria by the outgrowth of rhizoid filaments from the older parts of the shoots. These filaments grow downwards, creeping over the thallus, and form a kind of mantle over the lower parts of the shoots. Some of the Sphacelariacece are subject to the I 114 SEAWEEDS attack of parasitic entophytal Ckytridiacect, and thus present appearances which have been a source of error in their interpretation. The Reproductive Organs are both unilocular and plurilocular sporangia, and the different kinds occur as a rule on different plants. They are rarely (as in Batter sia) differentiated terminal cells of the axis or ordinary branches, but generally the terminal cells of special branches (their stalks, in fact), which arise sometimes singly, sometimes in tufts, in a con- siderable variety of relations to the axis and branches in the different genera. The unilocular sporangia are mostly round or oval in form, the plurilocular cylindrical or obovate. The latter in some instances may be branched at the base. The gemmse, which, so far as is known, are char- acteristic of this and the following order only among Phceophycece, are short branches, which cease to grow in length and send out two or three lateral short processes at the top, while the apical cell which had ceased to grow in length, emits a hair. The basal cell remains undivided and the gemma breaks off above it. On being set free, the terminal cells of the short processes or of the stalk grow out into a creeping filament, which bears new shoots as lateral branches. The Geographical Distribution is a general one, but possesses most representatives in north and south temperate seas, especially on the coasts of the North Atlantic and the Australian region. Battersia (peculiar to Britain), Sphacelaria, Chcetopteris, Clado- stephus, Halopteris, and Stypocaulon are represented 115 on the British coasts ; Sphacella in the Mediter- ranean ; Phloiocaulon, Anisocladus, and Ptilopogon in the Southern Ocean. Sphacelaria is represented in all seas. CHORISTOCARPACE.E. General Characters. — The thallus is filamentous, consisting of a single row of cells, and possesses an apical cell by which it grows in length. The repro- ductive organs, both unilocular and plurilocular spor- angia, arise as lateral outgrowths of the filaments, as do also the vegetative gemmae. The Thallus is in all cases filamentous and branched — never of more than one cell-row. The apical cell produces all the cells — i.e., there is no intercalary growth whatever. The Reproductive Organs. — The plurilocular sporan- gia of Discosporangium form a remarkable rectangular plate one layer thick, arising laterally from a slight outward bulging of the wall of one of the cells of the thallus. The loculi open outwards. In Chwisto- carpus the corresponding bodies more nearly approach those of Ectocarpacece. In Pleurocladia (a fresh-water form placed here) the plurilocular sporangia also resemble those of Ectocarpacece, and this genus further possesses unilocular sporangia (not known in Disco- sporangium and Choristocarpus) of similar type. The gemmae are known only in Choristocarpus and are lateral outgrowths of two, rarely three, cells, the upper being the larger. After they are shed the I 2 116 SEAWEEDS stalk-cell may proceed to form another, as in Sphace- lariacece. In vegetative propagation and in the growth of the thallus — viz., the absence of intercalary growth in length — the Chcristocarpacece appear to be related to the Sphacelariacece, while the type of re- productive organs points more to the Ectocarpacece. The Geographical Distribution of this small order, consisting of the three genera named (Pleurocladia , with two species, the others with one each) is very limited. Pleurocladia is purely fresh-water, one species occurring near* Berlin, the other in Kerguelen Land, while the other genera are both marine and confined to the Mediterranean. ECTOCARPACE.E. General Characters. — The thallus of the Ectocar- pacece is always of simple character and commonly consists of erect, simple or branched filaments spring- ing from a creeping filament or flat layer, or it may be reduced to a creeping filament or layer, from which the reproductive organs spring. These are both unilocular and plurilocular sporangia. The Thalhcs. — The primary creeping filament grows by division of the terminal cell or cells and bears in the most simple cases (e.g.. Streblonema) only reproductive organs and hyaline hairs with basal growth, or tufts of erect filaments (e.g., Ectocarpus), which grow by intercalary divisions without a definitely localised growing-point. These divisions are at first fairly general throughout the filament ; 117 they cease first at the apex, which becomes hyaline and elongate, and eventually continue only in places here and there. The filaments, simple or branched, are for the most part of one single cell- row and rarely divided by longitudinal walls. Some- times the cells of the erect filaments produce rhizoids which descending form a loose weft about the shoot. Where a basal horizontal layer occurs (e.g., Ascocydus and Phycocelis) it grows by peripheral cell-divisions and bears on its upper surface sporangia, hairs and paraphyses. The Reproductive, Organs. — The plurilocular spor- angia, which may be of one or more rows of loculi, are formed either by the differentiation of the upper por- tion of a branch or an intercalary part of it ; or they are definite outgrowths of cells of the erect filaments or primary basal layer. In Sorocarpus they occur in dense clusters ; but in the other genera they are free like the branches and mostly elongate, cylindrical, or oval in shape. They are commonly of several rows of loculi, but sometimes of a single row, wholly or in part. The gametes escape from all the loculi usually by an apical or lateral opening, and their conjuga- tion is of a peculiar character. In Ectocarpus silicu- losus, in which it has been carefully observed, the 9 gamete first comes to rest and is then surrounded by numerous ^ gametes, one of which succeeds in con- jugating with the $ gamete. That the $ gamete should first come to rest before becoming susceptible of conjugation recalls the case of CiMcria, and also in part that of Myriotrichia. The occurrence of bodies which can only be termed antheridia in 118 SEAWEEDS Eetocarpus secundus and E. Lcbelii, besides ordin- ary plurilocular sporangia, has already been men- tioned, and their full significance can be little more than guessed until much more is known of FIG. 29.— Ectocarpus coufervoide*. Filament with unilocular sporangia, and one plurilocular sporangium. Highly magnified. the reproductive processes in PhceopTiyccce. Mean- time, observation of the possible relation of their antherozoids, which fully resemble those of Fucacecc, Cutleriaccce and Tilopteridacccc, with the gametes of the plurilocular sporangia, may help the matter. PH^EOPHYCE^E 119 The unilocular sporangia occur (Pylaiella) in rows of intercalary cells or (Isthmoplea, Ectocarpus, &c.) as lateral outgrowths of the cells, generally sessile, and oval or globular in form. The relationship of the order is undoubtedly a close one with Sphacelariacece and other orders here placed near it, but facts briefly alluded to above compel one to look to orders like Tilopteridacece for other marks of near alliance. The Geographical Distribution of this large order is fairly general, but is most markedly abundant in the North Atlantic — though this probably means no more than that it has been most studied here. Streb- lonema, Edocarpus (many species), Isthmoplea, Pylai- ella, Sorocarpus, Ascocydus occur in British seas. SUB-CLASS II CELOROPHYCE^ THE ChloropJiycecc attain their finest development in fresh waters, but representatives of most of the orders occur in the sea. The multicellular forms never attain a higher stage of development than branched or simple cell-rows or flat expanses of indefinite form. The unicellular forms, however, in addition to the simple Algae usually so called, are represented by the multinucleate orders in which the highly developed thallus is a conspicuous feature. Oogamous reproduction occurs here, as in the Phceophycece, but mostly in fresh-water forms. It is represented in the sea in the marine species of Vaucheria only, and this type of oogamous reproduc- tion differs from that of Phceophycece (Fucacecc) in respect of the fact that the fertilised oosphere is not extruded but remains in situ in the parent plant. The fact of the occurrence of this high type of reproduction almost exclusively in fresh-water forms, and in those Algae with chromatophores most nearly resembling terrestrial vegetation, points to this group PLATE III. 1. COD1UM TOMENTOSUM. 2. CODIUM BUIISA. 3. HALICYSTIS OVALIS. 4. BPvYOPSIS PLUMOSA. Murray's Seaweed? Plate HI HijHey a«l.et UK TLanKiTi irap CHLOROPHYCE^ 121 as most probably nearer the lower limits of archegoniate plants than are any other Algae. The other Chloropliycece exhibit isogamous repro- duction— viz., the conjugation of equal gametes provided with cilia ; and in the case of the Con- jugatse, which are confined to fresh-water, the con- jugation of non-motile gametes. Non-sexual reproduction by zoospores and un- ciliated spores also occurs freely and abundantly in the Chlorophycece. ' The Caulerpacece, Vaucheriacece, Codiacece. Udo- teacece, Dasydadacece, and Valmiacecc which form a group together (usually called Siphonccc) are dis- tinguished from all other Algae by the fact that their often complicated thallus consists in reality of a single cell with many nuclei. In Valonia this one cell retains a primitive, more or less globular, shape, but in the other orders it is much branched and the branches gain coherence from being interwoven, laced together by haptera, incrustation, etc. In Ccmlerpa alone the lumen of the great cell is strengthened internally by numerous trabeculse or crossbeams that run from wall to wall. CAULERPACEJU. General Characters. — The order is represented by the single genus Caulcrpa (though systematists have proposed to split it up into several genera on wholly insufficient grounds) containing about eighty much varied species. They are exclusively marine. In some 122 SEAWEEDS respects Cauhrpa is the most singular type in the vegetable kingdom, since it attains a variety of habit rarely to be found within the limits of a very large order, though the plants are invariably composed of the branchings of a single cell. This cell obtains Fia. 30. — Caulerpa phyllaphlaston one-half natural size. (Ex. Fhyc.Mcm.) the mechanical support, elsewhere given by cross- walls, from a system of trabeculae or narrow beams, branching and traversing the cell from wall to wall in all directions. No reproductive organs are known, and the only method of propagation that has been discovered is by the separation of proliferous shoots. CHLOROPHYCE^ 123 The Thallus. — There is hardly any type of habit assumed by the higher terrestrial plants that does not find itself represented in this singular genus. There are species named from their resemblance to mosses, club-mosses, cacti, yews, etc., etc., while others are of extremely simple form. This differentiation is not con- fined to the green assimilating shoot, but extends to the root- system with its creeping rhi- zome-like extensions. The plants frequently attain con- siderable stature, and are in most cases of remarkable beauty. It has been said 1 that " Nature appears to have executed in the forms of this genus a tmtr de force in ex- hibiting the possibilities of the siphoneous thallus — in showing that it is possible for a unicellular organism to dis- play the varied beauties of outward form characteristic of highly-organised types ; to at- tain by means of a lattice- work of crossbeams within the cell-body that mechan- ical support effected by transverse septa and separ- ate, differentiated cellular structures for other Alga? and for the higher plants." 1 Trans. Linn. Soc. Bot., 2 ser., vol. iii., part 4, p. 207. 124 SEAWEEDS The erect, green, assimilating shoots are, as has been indicated, very variously branched, and the species are classified mainly on the characters so displayed. They root in fine sand or mud, and commonly possess a creeping surculus or rhizome which gives off roots below and erect foliar shoots at intervals after the fashion of higher plants. The network of trabeculaB or cross-beams is very dense in most species, and traverses the cell-contents in all directions. They spring from the substance of the outer wall-membrane (Fig. 32, c and d), and their principal function appears to be that of imparting support to the walls, though other functions have been speculatively ascribed to them without much show of reason. Dr. Correns has recently made a minute study l of the membrane, and has found that after treatment with sulphuric acid it has exhibited the formation of numerous sphasrocrystals, showing differences from those of cellulose demonstrated by Gilson and Blitschli. From the tests he has imposed, he has come to the conclusion that the membrane of Cauforpa does not consist of cellulose in the narrow sense, but of a substance not yet fully known, and different from callus, fungus-cellulose, reserve-cellulose, etc. He has obtained similar results from two species of Bryopsis, and is inclined from this to regard with favour the view, otherwise vaguely indicated, of a relationship of this singular genus with Bryopsis. So little indication is there, however, of relation- ship with other Alga3, that we have but the one fact 1 Ber. Deutsch. Bot. Gesdlsch., 1894, bd. xii., p. 355. CHLOROPHYCE.E 125 of its multinucleate character to guide us in assigning it a position here. The numerous species have been carefully examined without the discovery d c FIG. 32. — a, Caulerpa ligulata natural size ; 6, C. Holmesiana natural size ; c, transverse, and d, longitudinal section of stem, highly magnified. of any reproductive organs, which alone could show us its true position among the siphoneous Alga3. By an error there has been described a genus SEAWEEDS FIG. 33. — a and b, Caulerpa Carruthersii. c, section ; d, ideal figure of same half natural size. (Ex. Plnjc. Mem.) CHLOROPHYCE^E 127 Chlvrodictyon, said to be related to Caulerpa, and its description and figure have found a place in several important books of reference. It is however not even an Alga, but a lichen without the least structural resemblance to Caulerpa. The Geographical Distribution of the genus is throughout tropical and sub-tropical seas. There is one species (0. prolifera) in the Mediterranean, and several as far south as the Cape of Good Hope. The name Caulerpites has been given to a large number of fossil remains without, in a single instance, any reason other than a more or less vague resemblance of outward form. Since the genus itself resembles so many other types, this has been almost an inevitable error on the part of those palaeo- phytologists who choose to be content with mere impressions. A fossil Caulerpa (Fig. 33) from the oolite (Kimmeridge Clay), of which we possess casts in the round, stands on a firmer basis, since it not only exhibits resemblance of form, but occurs in the same beds with a marine fauna of a tropical climate. VAUCHERIACE.E. General Characters. — The order is represented by a single genus Vaucheria, and is the only one among the Siphonem with distinct oogamous reproduc- tion. Its simple, little differentiated thallus and highly developed type of reproductive organs, as well as the possession of non-sexual zoospores and a mode of vegetative propagation, mark it out from other 128 SEAWEEDS multinucleate plants. The species occur not only in the sea, but even more plentifully in fresh water and in places that are merely damp. The description of the natural order below is based on its general characters, and is inclusive of those derived from fresh-water species. The Thallus is simply a multinucleate filament, irregularly or dichotomously branched, and without farther differentiation. It forms much-branched, colourless holdfasts. The cell-wall is thin and the nuclei abundant in the protoplasm lining the walls. Sometimes crystals of oxalate of lime occur in the cell-sap. The chromatophores are oval and without a pyrenoid. Cross-walls occur only in connection with reproductive processes. The Reproductive Organs. — The oogonia and anthe- ridia are usually lateral outgrowths from a filament and occur side by side, though dioecious forms are known. The oogonia arise as round protuberances with a broad base, and gradually become more or less ovate and eventually cut off by a cross- wall at the base. The apex is generally papillate and the protoplasm becomes colourless here, and in this respect unlike the rest of the contents, which are coarsely granular and green, especially in the centre. The wall opens at the apex, and while the contents show a slight contraction, there is protruded a drop- like portion of a mucilaginous appearance. While this development is in progress the antheridium arises from the same filament as the oogonium and very near it. It is tubular in shape, and though sometimes straight is generally curved, and its cross- CHLOROPHYCE^ 129 wall is formed, as a rule, at the curve and not at the base. In V. synandra several occur together on what has been called an " androphore," while in V. de Baryana and V. piloboloides the antherozoids escape by lateral openings in the antheridium. There is, FIG. 34. — Vaucheria synandra. n, ti lament with oogonia and antheridia ; b, oogonium with antheridia; c, later stage of same; rf, escape of zoospore. Highly magnified. (After Woronin.) in fact, considerable variety, of use in classification, in the forms of the antheridia. The contents have very little colouring matter, and break up into a large number of small, biciliated antherozoids with the cilia pointing in opposite directions, which escape K 130 SEAWEEDS through the bursting of the apex. They gain access to the oosphere by its apical opening and impregnate it. The oospore thus formed invests itself with a wall, assumes a brownish colour, and generally under- goes a period of rest before germination. The plants that result from this germination are commonly reproduced by non-sexual means for several genera- tions of individuals before antheridia and oogonia are formed again. Zoosporangia, usually slightly club-shaped, are produced by the formation of a cross-wall near the end of a filament. The contents of the cell so formed, which are rich in green colouring matter, gradually contract into an oval shape, and escape by the bursting of the wall at the apex. There is thus normally but one zoospore in each zoosporangium, and it is large and ciliated at all points, except in some cases the posterior portion. These cilia are in pairs, and immediately beneath each pair there is a nucleus near the surface, as if the whole body represented an aggregate of zoospores which have failed to separate. In the formation of zoospores in other Qhlorophycece, there is usually one formed for each nucleus in the parent cell, and it is only natural to regard the case of Vaucherw, in the light indicated, as pointed out by Schmitz, who first observed it. It is more than a mere case of preserving the multi- nucleate character of the parent cell, since there is this definite relation of nuclei to pairs of cilia. It is of interest in this connection to note that during the escape of this great zoospore it sometimes gets nipped in two by the wall in passing the opening, and CHLOROPHYCE.E 131 each portion becomes a zoospore. The zoospore retains the character of its parent cell in respect of a cell-sap cavity, traversed by threads of protoplasm. A certain resemblance of this remarkable body to the Volvocinece (fresh water) suggests inevitable speculations which may easily be made too much of. In the allied genus Derbesia (Codiacece) there are formed a number of zoospores in each zoosporangium —but before this the parent cell contains a large number of nuclei which unite with each other, and thus become reduced to a number equivalent to the number of zoospores. The zoospores soon lose their cilia and settle down, becoming invested with a cell-membrane. They do not rest, however, more than a day or two at most, and germinate by the emission of one or more tubes. It happens in some species that antheridia and oogonia are occasionally formed on these filaments immediately after germination. Motionless spores are produced by the formation of cross-walls near the apex and the abstriction of this portion, which first swells into an oval or globular form. It secretes a new cell-wall, and is set free by the dissolution of the original wall. In some cases such spores germinate soon, in most they rest before germinating. This spore-formation, like the more extensive formation of such bodies by segmenta- tion of the thallus, is often caused by injury or unfavourable external conditions, and is more charac- teristic of the species inhabiting fresh water, which are moreover subject to the attack of rotifers (Notommata) giving rise to galls. K 2 132 SEAWEEDS The Geographical Distribution is general, and the best-known British marine forms are V. dichotoma, V. synandra, V. Thureti, V. sphcerospora, and V. litorea. CODIACE^E. General Characters. — The order is represented by the genera Codium and Bryopsis, differing in habit and mode of branching, but agreeing in the production a FIG. 35. — Codium tomentosum. a, part of section showing club-shaped cells of periphery with lateral sporangia highly magnified ; b, one such cell with sporangium more highly magnified. of gametes of two sorts, between which conjugation, though it has not been observed, may yet be pre- sumed with much probability to occur. Derbesia, also included here, is of more simple vegetative develop- ment, and possesses, so far as is known, but one kind CHLOROPHYCE^ 133 of reproduction, viz. by zoospores. It is consequently an aberrant type, but its inclusion may be justified on grounds dealt with later. The Thcdlus. — The beautiful feather-like fronds of Bryopsis recall the habit of certain species of Caulerpa, but between the genera there is little more than this superficial resemblance. Wille has stated that in old stalks of Bryopsis trabeculae sometimes occur, but the observation is open to question. If established, it would certainly help in the elucidation of the position of Caulerpa, especially when taken together with the observation of Correns (p. 124) on the character of the cell-membrane. The thallus of Bryopsis consists of a single branching cell differ- entiated into rhizoids and erect shoots. The shoots consist each of a primary stalk, which either itself bears the ultimate lateral branches (which for con- venience may be called leaves with apologies to formal morphology) or it may first branch repeatedly. The variations in this respect and in the occurrence of the leaves, whether opposite, alternate, secund, or in irregular spirals round the stalk, are characters of use in the discrimination of the species, though they are to be used with caution, since variation occurs within certain limits. The erect feathery fronds of a beautiful deep green are not only very graceful, but furnish objects of much interest for the microscopical study of the cell. In the middle of the cell is a vacuole, and the protoplasm lining the wall contains numerous nuclei and oval, flat chromatophores, each with one pyrenoid. Bryopsis, like Derbesia, Vaucheria, and Caulerpa, has all its filaments free, and not 134 SEAWEEDS interwoven like so many other multinucleate Algae. In the thallus of Codium two layers may be distinguished, though their component elements are all branches of the same cell. In the middle there is a densely interwoven mass of filaments which end outwards in club- shaped apices arranged perpen- dicularly to the surface. These club-shaped ends lie parallel to each other, and form a kind of pallisade tissue (Fig. 35). The spongy thallus so formed is of varying habit, either elongate and dichotomously branched (C. tomentosum), or in globular cush- ions, or flat and encrusting, with- out definite stalk and with very slightly developed rhizoids. The lumen of the cell is sometimes interrupted (apart from the formation of reproductive bodies) by plug-like thickenings of the membrane. In Dcrlesia, the single fila- ments of the thallus show no differentiation of frond. They are dichotomously but sparingly ,. branched, and are rarely, but : with no regularity, septate in CHLOROPHYCE^E 135 the older parts. The chromatophores are oval, and in some species contain a pyrenoid, in others not. The Reproductive Organs. — In Bryopsis the gametes are produced on different plants within the lateral leaves, which undergo no modification to this end beyond the separation by a cross- wall or stopper from the stalk. They are of two sorts ; the smaller male gametes being elongate and orange-coloured except at the anterior end, the female much larger and green, with a red spot between the green contents and the colourless portion. They are both biciliate. Sometimes the female gametes have a spike-like projection from the posterior end. In C i odium the gametes are borne in lateral ovate sporangia (game- tangia) arising from the club-shaped pallisade cells and lying among them, but never projecting beyond them (Fig. 35&). They are separated at the base by a cross-wall or stopper. The gametes are here also of two sorts, large green ones (female) and smaller yellow ones (male), both biciliate. Conjugation has not been observed in either genus, but is extremely probable from the fact that Berthold's experiments in cultivating the green gametes were never successful, unless when these were mixed with the smaller ones. Analogy points also to this interpretation of their character. If this be so, we possess in the Codiacece a valuable link between Vaucheriacece and the other multinucleate orders. Codiacece would thus occupy to Vaucheriacece much the same position as Cutleriacece to Fucacece among the olive-brown seaweeds. As for non-sexual 136 SEAWEEDS reproduction, however, the parts are reversed, since none is known in Bryopsis and Codium. Went,1 however, in studying Codium tomentosum, found both kinds of zoosporangia on the same individual plants. Sometimes the small gametes do not appear until after the dispersal of the larger ones — and they may be formed from the same basal cell. He never observed any union of the two sorts, and successfully cultivated the contents of the larger kind of gametangia or sporangia. But it is im- possible to say in this case whether there had been conjugation or not. In Derlesia we have only non-sexual reproduction, and, if the genus be rightly placed here, this fact also may have its significance, since Derbesia in other respects strongly resembles Vaucheriacecc. The zoosporangia arise as lateral outgrowths from the filaments, and contain 8 — 20 zoospores with a fringe of cilia on the anterior end. As mentioned under Vaucheriacece the nuclei in the parent cell become reduced by union to the number of zoospores formed. However, it is plainly an aberrant type, and needs further investigation. The Geographical Distribution. — Bryopsis occurs in all seas, but more abundantly in warmer waters. Codium also has an extensive range in the warm and temperate seas of the world. Derbesia belongs to the north temperate and tropical Atlantic (both Euro- pean and American), and occurs also on the Australian coasts. All three genera are British, and (with Vaucheria and Halicystis) constitute our represent- 1 Vergrad. d. Ned. Botan. Vereeniging, 1889. CHLOROPHYCE^ 137 atives of multinucleate plants. The British species are Bryopsis hypnoides and B. plumosa, C odium tomentosum, C. Bursa and 0. adhcerens, and Derbesia tenuissima. UDOTEACEAE. General Characters. - - The reproductive bodies (known only in the case of one genus, Halimeda) are zoospores produced within zoosporangia. It is not known whether they are zoospores or gametes in fact, since no observation has been made of their conjugation. The principal interest in the order is to be found in the vegetative part of the thallus, since the generic forms are not only of striking out- ward appearance, but of singular structure. The cohesion of the filaments composing the fronds is attained by their being interwoven (Avrainvillea), by incrustation of carbonate of lime (Penicillus and Halimeda), or by incrustation more or less partial with the addition of lateral haptera or holdfasts binding the filaments together ( Udotea). The Thallus. — In Avrainvillea ( = Fradelia, Chloro- plegma, Ehipilia, Chlorodesmis) the frond is of simple structure, consisting of unicellular filaments, re- peatedly branched, and interwoven so as to form a stalked or sessile, fan-shaped, felt-like frond above and a dense mass of rhizoids below. The fan-like fronds are tough and spongy, and often lacerated at the edges. The filaments are dichotomously branched, and each branch is constricted at the base. In some of the species, the whole of the filament is constricted 138 SEAWEEDS at short intervals, so as to resemble a chain of beads, though the constriction is not so deep as that would imply. This basal constriction is found also in Penicillus, and some species of Udotca. At the basal constriction in A. comosa there are sometimes formed 9 f PIG. 37. — a, Arrainvillea longicaulis half natural size; b, frond filament of A. Mazei ; c, ditto of A. lonoicanlis ; d, ditto, showing beginnings of branches ; e, filament of A. comosa, with stoppers ; g, tip of filament of A. papuana ; f, the same with dark brown contents, b, e, d, e, f, g, highly magnified. stoppers like those of Bryopsis and C odium. In A. com.osa the filaments of the frond are free, and in young specimens of A. papuana they are very little interwoven. The cell-wall is uniformly thin except in A. longicaulis, where it becomes (in the rhizoids) so much thickened in places as to obstruct the lumen CHLOROPHYCEJE 139 of the cell. The chromatophores are rounded or polygonal, always with a clear central pyrenoid. Starch grains are very numerous, especially in the rhizoids, and are kidney-shaped in A. papuana and spindle-shaped in A. comosa. A yellowish or brownish colouring-matter is very abundant in the protoplasm and occurs very densely at the tips of the filaments, where it presents a dark resinous appearance. The rhizoids are less regularly dichotomous in their branching, and in most of the species enclose among the filaments masses of coral-sand, small shells, &c., so much so that the rhizoids of A. papuana, when drawn forth from the mud or coral-sand in which the plant grows, present the appearance of a cylindrical mass of crumbling mortar. In A. longicaulis there are formed rhizome -like, creeping bodies that con- nect large numbers of fronds. They are round like the stalks, and of the thickness of a finger. In this way so many plants are connected — all of them ramifications of a single multinucleate cell — that probably there is no parallel to it to be found in the plant world in respect of its dimensions. Penicillus agrees in many essential respects of minute structure with Avrainvillea, such as the dichotomy of the filaments, but the plants do not outwardly resemble each other. The rhizoids are much alike, but the stalk is thickly incrusted with carbonate of lime, while from its summit the frond filaments wave free. Sometimes they are given off singly here and there from the surface of the stalk below the summit, but generally in a dense apical mass like a mop. In some species the filaments are 140 SEAWEEDS free or almost free from incrustation, while in others each filament is thickly incrusted. In the latter case they are constricted at short intervals like a string of beads. Penicillus mediterraneus has a growth- form formerly known as Espcra mediterranca, in which there is no stalk, but all the filaments are free. Similarly the species of Avrainvillea with free Fia. 38.— Penicillus capitatus natural size. filaments and no stalk (A. comosa) were formerly placed in a separate genus Chlorodesmis. A fossil genus Ovulites from the Eocene (Paris basin) has been placed here by Munier Chalmas, but though there is no doubt it is a calcified siphoneous Alga and does riot belong to the Foraminifera as formerly thought, its position near Penicillus is not so certain. CHLOROPHYCE.E 141 It might, with almost equal fitness, be placed else- where. Its remains are in the form of small egg- shaped bodies with a hole at each end, and each of these is taken by Munier Chalmas to represent the beads in the filaments of a Penicillus-like plant. In Rhipocepliahis there is an erect incrusted stalk giving off at intervals numerous small fronds also incrusted, composed of dichotomous filaments. It forms a transition to the next genus. Udotea has the same fan-like, stalked fronds as Avrainvillea, but in this case they are in- crusted, in some species very little, in others thickly. The filaments are little interwoven, but in addition to the incrusta- tion they are bound together by numerous short lateral branches terminating in hap- tera or sucker-like holdfasts. The fronds in some species are _ .„ „ , f - . FIG. 39. — Rkipocfpkalw beautifully Zoned, and there IS fhamix one-half the natural in nearly all a tendency, more or less marked, to proliferation at the margin. The round bodies, figured by Kiitzing, which have been taken to be zoosporangia, are probably of foreign origin. Halimeda is the most singular of the group in the 142 SEAWEEDS form of its fronds. These are chains of incrusted segments jointed together cactus-fashion, and the shape of the segments varies with the different species, but is usually more or less heart- or kidney-shaped or irregularly round. These chains usually arise from cylindrical incrusted stalks, and, i ' FIG. 40. — a, Udotea Favonia half natural size ; b, the same in longitudinal section highly magnified. like the other genera, are firmly rooted by a mass of rhizoids. The filaments composing the fronds are dichotomously branched, but of irregular form. The central ones are large and elongated in the direction of the axis, while from them spring shorter ones passing outwards towards the margin, forming a kind of cortical zone (Fig. 416). CHLOROPHYCE^E 143 FIG. 41. — a. Halimeda monilis half natural size ; b, the same in longitudinal section highly magnified. Callipsygma is a genus of which only an imperfect description has been published. It appears to be related to Avrainvillca and Udotea. A fossil Alga of Devonian Age, Nematophycus, has 144 SEAWEEDS been placed in this group by Mr. Carruthers. It possesses stalks more than a foot in circumference, and must have been a colossal member of the group. Its filaments are entirely free from cross-walls, and are bound together by very fine lateral branches, though these do not appear to terminate in haptera, but rather wind round the larger main filaments. The Reproductive Organs are known only in the genus Halimeda, and are zoosporangia producing biciliated zoospores. The zoosporangia arise from the margins of the flat fronds, and are round or pear-shaped bodies borne on branching filaments. They are green in colour and not separated by cross-walls from the filaments of the thallus. The zoospores are very small and green in the posterior part, hyaline at the ciliated end. They are pro- duced in great numbers. No observation has been made of their possible conjugation or germination. The genus Codiophyllum, placed here doubtfully by Wille, is identical with a red seaweed Tham- noclonium. The Geographical Distribution. — The order is almost wholly tropical, though Penicillus, Udotea, and Halimeda have each one representative in the Mediterranean. All the genera except Callipsygma- (of somewhat doubtful validity) are represented in the West Indies, while Avrainvillea, Halimeda, and Udotea are abundant also in the warm Indian Ocean, Malay Archipelago, and Pacific islands. Penicillus has not quite such a wide range, but occurs in the West Indies, Australia, and Moluccas. Callipsygma is known only from Australia. CHLOROPHYCE.E 145 DASYCLADACE.E. General Characters. — The thallus consists of an erect axial cell with no cross-walls, bearing whorls of lateral branches, some hair-like and sterile, others of definite growth and fertile. The shoot is attached by rhizoids to its substratum. Reproduction is effected after the conjugation of gametes. The order is naturally divided into at least two sub-orders, viz. Acetabulariece and Dasycladece. Acetabulariece. The Thalhis.—The erect axis bears at its summit a disc-like cap, consisting of radiating chambers, which in the species of the genus Acetabularia (as pIG- 49. Acicularla Schenckii natural size. (After Solms-Laubach.) formerly denned) are united to each other laterally so as to form a firm disc. Other species were for- merly placed under Polyphysa, but that genus is now 146 SEAWEEDS joined to Acetdbularia, and these are characterised by the radiating chambers being free from each other, and by other differences. The best studied species is A. mcditerranea, and a FIG. 43. — a, Acetnbularia exigua decalcified (magnified) ; b, Acicularia MiJbii, longitudinal section with two superposed caps ; c, Acetabularia Calyculus, cap-ray with spores showing upper and lower coronse ; d, A. exigua, spores with lid. b, c, d highly magnified. (After Sohns-Laubach.) description of it will enable us to understand the variations displayed by the others. It has an erect stalk six to nine centimeters high, with a cap about a centimeter across, strongly calcined and parti- tioned into a number of regular radial chambers, CHLOROPHYCE^E 147 which unite in the centre above the insertion of the stalk. This central area immediately above the stalk is covered with a flat circular membrane above, and is surrounded, where the radial chambers are given off, by a continuous circular cushion (Fig. 43c), the corona superior. This corona consists of as many segments as there are radial chambers, and each segment bears the scars of hairs that have fallen off or remain incompletely developed. In like fashion below the marginal radial chambers there is another cushion, the coi^ona inferior (Fig. 43c), corre- sponding with the upper one, but bearing no hair- scars. A lower but less- marked cushion exists, but it gradually becomes merged in the central area. Each marginal ray stands in communication with the relative segments of the corona (superior and in- ferior), and these in turn are marked off from the central area by a fold of the membrane with a central opening, which frequently becomes closed by subsequent thickening. The radiating chambers bear the spores (gametangia). In the species formerly placed under PolypJiysa (which has been merged into Acetabularia by Graf zu Solms-Laubach), the sporangial rays are com- pletely free from each other, and not united into a firm disc, as in A. mediterranea. They are quite un- calcified in some species and very slightly so in others and have no corona inferior, while the corona superior is represented by free knobs bearing hairs. A. mediterranea takes several years to attain the formation of a fertile cap. In the first season no cap is produced, but merely erect stalks, with occasion- L 2 148 SEAWEEDS ally irregular protuberances at the tip. The plant dies down, and the lower part of the stalk, closed with a membrane at the base, remains alive during the winter, and may be described as consisting of two portions, viz. the foot, calcined and irregularly branched, and a basal, rhizoid body. This basal body increases with age, and acts as a storehouse of reserve-material for further growth. In the following year a cap without spores is produced, while the stalk, before this happens, gives rise to one or more whorls of branched hairs. These hairs are not calcined, and are soon thrown off, leaving only rings of scars on the stalk of the cap-bearing plant. After giving rise to several sterile plants in succes- sion, eventually a fertile cap is borne. This production of whorls of hairs is interesting, not only in throwing light on the homology of the sporangial rays, but in relation to the neighbouring genus Halicorync, which possesses alternate whorls of fertile and sterile branches. The sterile whorls consist of repeatedly multisect hair-tufts developed in groups of eight ; but they soon fall off, and leave round scars on the stalk. Between these on the full- grown plant there are sixteen-branched whorls of a different kind — the sporangial rays, which are com- pletely free. Each ray is furnished on its upper sur- face near the base with a small protuberance bearing one or two diminutive rudimentary hairs, recalling the corona superior. Interesting also in this con- nection is the fact that Acetabularia crenulata, in the normal course of its development, and not as a monstrosity, produces several caps in succession CHLOROPHYCE^] 149 above each other (Fig. 44). Halicoryne is, in fact, a better representative of the Acetabulariece than the disc-forming Acetabularia, since it supplies us with a link to the other Dasycladacew, from which it is distin- guished, however, by the alternation of fertile and FIG. 44. — Acetabularia crenulata with several caps in longitudinal section highly magnified. (After Solms-Laubach.) sterile whorls, in addition to the more important characters of the sporangia. The genus Chalmasia (like the former genus Polyphysd) has no corona in- ferior, and while closely resembling Acetabularia in most respects, agrees completely with Halicoryne in the structure of the spores. In Halicoryne their mem- 150 SEAWEEDS branes are strongly calcified and stratified. Acicu- laria, which has both corona superior and inferior, is distinguished from Acctabularia by the fact that its spores are strongly calcified (those of Acetabularia are free from incrustation), and adhere in clusters. Acicularia contains both living and fossil forms. The latter are A. Andrussowi and A. miocenica from the Miocene and A. pavantina from the Eocene. The Reproductive Organs. — The life-history has been followed in great detail in A. mediterranea. Within the marginal chambers the spores (game- tangia) are produced in considerable numbers. They rest from one to three months after extrusion, and then opening by a lid at one end emit the gam- etes. The gametes are biciliated, of equal size and same shape, and with a red spot. They conjugate usually in pairs, and not only so, but in fours or fives, sometimes laterally sometimes in reversed position ; but conjugation takes place only between gametes derived from different gametangia. After conjuga- tion they do not immediately settle down, but remain in a state of movement with the cilia in pairs still vibrating. The zygote, so formed, eventually gives rise to a new cap-bearing generation, after a rest of about five months. If we apply the theory of alternation of genera- tions to the life-history of Acctabularia, we must regard the product of the zygote, viz. the cap-bearing plant, as the non-sexual one, terminating in the pro- duction of spores within its radial chambers. The sexual generation is represented by the spores which without any vegetative manifestation become them. CHLOROPHYCEJE 151 selves gametangia, bearing gametes which conjugate, and so form a zygote. The Geographical Distribution. — The Acetabulariece are a tropical and subtropical order. Of the fifteen species of Acetdbularia, one however occurs in the Mediterranean and A. Peniculus on the West Australian coast. They occur in all tropical seas. Halicwyne has two species, one (H. Wrightii) in the China sea and the other (H. spicata) on the shores of New Caledonia. Chalmasia has only one species, G. Antillana, and Acicularia is represented alive only in the warm Atlantic (West Indies and South America). Dasycladece. The Thallus. — This sub-order is characterised by the persistence of the whorls of hairs and by there being no distinction between fertile and sterile whorls — they are all fertile, except locally in the joints of Cymopolia — and by the gametangia being terminal members of the lateral branches, except in Botryophora where they are of lateral origin on these branches. In Dasycladus the shoot is nowhere incrusted, and consists of a short erect axial cell without con- strictions or cross-walls, attached to the substratum by a holdfast, and clothed with whorls of branches, about twelve to each whorl. These are again branched in whorls several times, each successive whorl diminishing outwards in length and in the number of branches. At the apex of the ultimate whorl there is produced a single globular sporangium 152 SEAWEEDS (gametangium) on a short stalk surrounded by the end branches. In Botryoplwra there is also no incrustation, and the habit is similar but more lax, the plants being larger and the branches not so densely packed together. The sporangia are lateral and occur several together on the base of the branches. They contain a large number of spores with membranes, which may be gametangia, but no observation of their germination has been made. There is only one species, which was originally described under Dasydadus, and very probably that will prove to be its true position. Neomeris has an incrusted shoot of otherwise similar character to Dasydadus. The lateral branches bifurcate and end each in a swollen tip. These tips are arranged in rows of facets externally, forming a sort of outer cortex to the shoot. At the bifurcation of the lateral branches and between them, occupying morphologically the summit of the original lateral branch, there is a single oval sporangium, with one spore as to the development of which nothing is known. Bornetella differs from Neomeris in the branches being in two or three successive whorls and in the sporangia being of lateral origin, and containing several spores. Cymopolia is of very different habit since the main shoot, the axial cell, is repeatedly branched dichoto- mously. It consists of chains of cylindrical, incrusted portions, separated from each other by short, not incrusted, joints in which the branching has its origin. The apex of each branch has a terminal tuft of hairs like a brush. At the incrusted portions FIG. 45.—a,Neomeri8 annulata magnified ; 6, Bornetella oligospora magnified (After Soluis-Laubach.) 154 SEAWEEDS there are borne the fertile whorls, each with a secondary whorl and a terminal sporangium in the middle. The ultimate whorl of these lateral branches, overtopping the sporangia, form a kind of cortex (as in the last genera), but without definite facets externally. The space between them and the axial cell is originally of slimy consistence, but becomes the seat of incrustation. At the joints there are no fertile whorls and the lateral branches are here simple and decrease in length from below upwards. It is a similar formation which bears the terminal tuft of hairs. No observation of germination has been made. A number of fossil genera, such as Decaisnella, Haploporella, Dactylopom, Dactyloporella, Uteria, and Polytrypa from the Eocene, and Munieria, Gh/roporella and Triploporella of Cretaceous age are certainly nearly related to the forms just described. The Reproductive Organs. — Dasycladus is the only genus in which the gametes have been observed. We have seen that in Acetabularia the gametes unite only with those from other gametangia, but in Dasycladus the matter is carried a step farther, since the gametes are here incapable of conjugation with others from the same plant, and indeed they conju- gate only with those of particular plants, not with any other indiscriminately. It appears from this obser- vation that though these gametes are apparently all alike, there yet resides within them a definite character indicating a difference of sex, though this is not determinable by us from their structure. Judged by itself alone, Dasycladus appears to pre- CHLOROPHYCE^E 155 sent us with a case of a spore (gametangium) borne not in a sporangium as in Acetabularia, but free on its branch. One may regard this spore, however, in the light of a sporangium which omits to form spores and itself gives rise directly to gametes, and this view is much strengthened by comparison with the case of Botryophora and Bormtella, where there is an actual sporangium, with spores of similar origin, though we do not know from observation what the fate of these may be. Dasycladus by itself appears to exhibit perpetual production of sexual generations, while in Accta- bnlaria it is this generation that remains unde- veloped, and is represented only by the spores (gametangia). Such interpretations, however, must remain purely speculative until further light is thrown on the matter by an examination of the life- history of such forms as Botryophora (so near Dasydadus), Neomeris, Bometella, and Cymopolia. The Geographical Distribution, like that of Aceta- bulariece, is in tropical and subtropical seas. Dasy- cladus has one species occurring in the Mediterranean and the Canary Islands. Botryophora also has one species confined to the West Indian region. Neomeris has several species, one in the West Indies, others in Madagascar, the Malay Archipelago, and the tropical Pacific. Bornetella is Australian, and occurs also in the tropical Pacific. Cymopolia barbata occurs in the West Indies, the Canary Islands, and has been re- corded from Cadiz, but not from the Mediterranean, while C. van Bossei represents the genus in the Malay Archipelago. 156 SEAWEEDS VALONIACE.E. General Characters. — There is scarcely any Order to which it is more difficult to assign distinctive general characters, owing to the much varied structure of the vegetative organs and our ignorance of the reproductive processes in most of the genera. The thallus ranges in variety from a single large cell with rhizoids up to forms of complex structure with stalk and frond. The cells of the thallus are fre- quently linked together by haptera or holdfasts. The reproduction by zoospores described for Microdictyon and Anadyomene certainly needs minute re-investi- gation, but that of Siphonocladus and Valonia stands on a firmer basis of observation. In Valonia spores with cell-membranes arise by free-cell-formation within the great lumen of the cell, but their ger- mination has not been observed. The following types indicate the character of the thallus, and ex- hibit a series connecting the multinucleate Algae with the other Chlorophycece, or at all events pointing to such a connection. The Thallus. — The most simple type is that of Valonia ventricosa (Fig. 46a), which consists of a single cell, generally varying in size from that of a goose- berry to a hen's egg. This enormous cell, much the largest cell of isodiametric shape known to us, is attached to its substratum by rhizoids at the base, and presents a uniform green appearance, except on the cessation of its vegetative life, when the protoplasm with the chromatophores, which had lined the wall, CHLOROPHYCE^E 157 shrinks into the great lumen of the cell -sap cavity and leaves the plant as a translucent sphere. In this condition it frequently parts company with its attachment and floating to the surface is drifted ashore. It is common in the West Indies, and reaches Bermuda, where the plants are often drifted ashore in this translucent state, and are called " sea-bottles " by the inhabitants. Halicystis ovalis, which resembles this plant in shape, but is much smaller, occurs in the Clyde Sea area (and extends from western France to the Faroes and Scan- dinavia). Its systematic position is uncertain, since we know nothing of its reproduction, but so far as may be judged by the structure of its membrane, which shows none of the striation and very little of the stratification of Valonia, by its chromatophores, which have no pyrenoid, while those of Valonia possess one in many cases at least, and by the substitution of a sucker-like disc for rhizoids, it must be placed apart from Valonia. Schmitz suggests that its vegetative structure recalls the freshwater Botrydium most closely, and is mainly distinguished from it by the absence of rhizoids. However, this is true also of a comparison with Valonia, and until we know its reproduction any speculation must carry little weight. In other species of Valonia there occurs a remark- able form of branching, if it may be so termed. Small portions of protoplasm and chlorophyll gather opposite more or less definite parts of the mem- brane, generally near the apex, and separate them- selves from the rest of the contents by the formation 158 SEAWEEDS FIG. 4(5.— a, Vah-nia ventricosa slightly reduced ; I, section of wall ; c, d, e, reproductive bodies ; /, chromatophores 6, c, d, e, f, highly magnified ; g, Htilicysti* oralis natural size ; h, section of wall ; fc, /, chromatophores h, k, I, highly magnified. (Ex. Phyc. Mem.) CHLOROPHYCE.E 159 of a membrane of the form of a watch-glass. This cell then proceeds to grow out from its parent cell, and assumes a similar shape. The process is then repeated, and *by this means a thallus is produced, generally with irregular branching, but sometimes having the appearance of successive dichotomous or verticillate branching from the apex of each generation of cells. Dictyosphceria favulosa in its early state is an irregularly globular mass of large cells (Fig. 47c), the interior of the mass becoming hollow with its growth. It eventually bursts, and the thallus is then ir- regularly lobed. It consists of numerous cells in several layers, all of these being bound together by sucker-like holdfasts, short where the cells are closely packed, and long-stalked where they are more loosely aggregated. There are remarkable internal spines projecting from the cell-membrane into the cell- cavity (Fig. 47 A). D. scricea has a different arrange- ment of holdfasts, but the main points of structure are the same. Wille describes the origination of the cells of Dictyosphceria within a mother cell. This may be the case, but an examination of many early stages of D. favulosa does not bear it out. However, it would be in harmony with what is known of Valonia. The placing of Elastophysa among the Valoniacece is uncertain. The plants are green, very irregular, much lobed cells with long colourless hairs. It is certainly multinucleate, but its reproduction is un- known, and its Vegetative characters inadequate for determining its true position. Siphonocladus is a simple, minute, multicellular 160 SEAWEEDS y a d FIG. 47. — Dictyogpha-ria favulosa. a, b, <•, young specimens ; d, diagram of cells joined by haptera ; e, haptera in side and surface views ; /, D. sericea, surface view of frond ; g, bordering cells with haptera ; h, internal projections from cell-walls of D. favulosa. Variously magnified. (Ex. Phyc. Mem). branched type, which leads on to forms with a higher differentiation of thallus. In Apjohnia there is a main stalk with numerous rugose constrictions and CHLOROPHYCE^ 161 with no cross-walls, giving off rhizoids below and dichotomous branches above. Chamacdoris has a similar stalk, with its branches given off in a great terminal tuft — in habit like Penicillus in this respect. It appears very probable that the stalk is persistent and renews its crop of branches, both in this genus and at all events in some of the species of Struvea. The large species of Struvea are among the most beautiful of Algae. At the summit of the long rugose stalk without cross- walls there is borne a flat frond, through which the stalk is prolonged as a mid-rib. This mid-rib gives off opposite branches, which are again pinnately branched, and in some species these are similarly branched again and again. Where these pinnae meet they are all bound by sucker-like haptera (Fig. 48e), and the frond presents the appearance of a lovely piece of lace. S.plwmosa, S. macrophylla, and S. pidcherrima are the largest and finest species. Only three specimens of S. macro- phylla have been found, two of them being in the herbarium of Trinity College, Dublin, and one in the British Museum. S. pulcJicTrima is even more rare, one specimen, not quite complete, being in the British Museum, and a fragment in the Edinburgh herbarium collected at the same time. The more slender forms have stalks unmarked by rugosities. The forms described as species of Spongocladia were long puzzling. They are dense wefts of interwoven filaments, with walls so much thickened in places as to obliterate the cell-lumen. They grow in intimate association (Symbiosis) with sponges, and assume to some extent the habit of these animals. It has M 162 SEAWEEDS Fio. 48. — a. Slrnven macrnpltylla slightly reduced ; b. .9.— Catenella opuntia. a, section of fertile part of thallus showing numerous trichogynes ; ft, carpogonium with trichogyne highly ina°Tiified (After Harvey Gibson). its trichogyne, most of which are fertilised. The gonimoblast forms chains of carpospores which escape, since there is no ostiole, after rupture of the cortex "already weakened by the numerous aper- tures occurring in it through which the trichogynia pass to the exterior," as Prof. Harvey Gibson has described. 222 SEAWEEDS The Geographical Distribution of the order is throughout all seas. The British genera have already been indicated, and the others, too numerous to particularise here as regards distribution, are some of them local, but most with a fairly wide range. The British forms have mostly a wide distribution else- where in the ocean. RHODYMENIACE.E. General Characters. — More Rhodophycew probably conform to this type than to any other, and it is sub- divided into a number of important families. The carpogonial branches and the parent cells of the auxiliary cells are developed in close proximity and form definite procarpia. After fertilisation has been effected the auxiliary cell is developed, and with it the fertilised carpogonium enters into conjugation by means of a very short ooblastema filament, the auxiliary cell sometimes sending forth a short process to meet it. Usually, and this is true of most Rhodophycecc, the carpogonial branch is so curved that the carpogonium is brought into close vicinity with the auxiliary cell, and the ooblastema filament is therefore either very short or may be suppressed owing to the actual contact of these bodies. The auxiliary cell after this conjugation sends out the gonimoblasts towards the outside of the thallus, and the following families are established mainly on variations in this process, and its results in the form of cystocarpic fruits. PLATE VII. 1. CERAMIUM DIAPHANTJM. 2. RHODYMENIA PALMATA. 3. DELESSERIA SANGTJINEA, 4. POLYSIPIIONIA BRODIAEI ^. LOMENTARIA ARTICULATA. Plate VII Berjeau^C KigHey ."Wet iith Kanhart RHODOPHYCE^, OR FLORIDE^E 223 Sphcerococcece. Nearly all the members of this family have much branched fronds without foliar expansions, the branches being frequently robust and of dense texture. The gonimoblast is formed within the FIG. 70. — a, Phacelocarpns Labillardierii natural size ; 6, cystocarp of Gracilaria confervoides highly magnified. (After Thuret and Bornet.) thallus, and produces its cystocarp within a wall formed of the peripheral thallus tissue which arches outwards and has an opening at the apex for the escape of the carpospores. The fruits consequently look like semiglobular swellings on the branches. Within this fruit-cavity and at its base there is formed a placenta frOm which the spores are pro- duced, and its surface is free from the over-arching 224 SEAWEEDS wall. The gonimoblast itself is a much branched tuft of closely packed filaments forming this convex placenta, and the carpospores are produced singly or in series on the points of the filaments at the free surface of the placenta within the cystocarp. The Geographical Distribution is a very wide one and the forms are more abundant in the warmer seas, especially the two large genera Gracilaria and Hypnea. The other notable genera (Spkcerococcus, Corallopsis, Phacelocarpus, and Calliblepharis) of the family have for the most part a similar distribution but are much more local. Sphcerococcus, Gracilaria, and Calliblepharis are represented in British seas. Gracilaria lichenoides furnishes the substance called " Ceylon moss " from which at least one kind of Agar-Agar is prepared, a substance of much use in the cultivation of Bacteria. Rhodymeniece. As in other families, so in this one, there is con- siderable diversity in the vegetative characters. Rhodymenia, of which R. palmata yields the edible " dulse " of our coasts, and Plocamium have fronds of conspicuous size, beauty, and firmness of texture, while such genera as Chylocladia, Champia, and Lomentaria with their hollow, tubular, and jointed fronds represent a simpler type. Chylocladia is well adapted to exhibit the typical constitution of the thallus of Rhodophycece built up of separate filaments, as Schmitz has established. By boiling a specimen in distilled water and afterwards pressing it under RHODOPHYCE^, OR FLORIDA 225 the cover-glass the points of the branches may be resolved into their component filaments, each with its apical cell and each exhibiting its history of division. All three genera (Champia, Chylocladia, and Lomentaria) have hollow tubular fronds filled with a gelatinous substance which also coats the FIG. ^l.—a,Chylocladia kaliformis, carpogonial branch with trichogyne ; b, fusion of cells of carpogonial branch to form one cell ; c, the fertilised auxiliary cell with the two nuclei near each other ; d, optical section of young fruit highly magnifled. (After Hauptfleisch.) outer wall. The first two have diaphragms inter- rupting the continuity of this tube, the last has none. With various minor modifications their repro- ductive processes, which have been studied in great detail by Hauptfleisch, arc essentially the same and typical of the family. In Chylocladia kaliformis the Q 226 SEAWEEDS carpogonial branch consists of four cells, and is formed, near the apex of a growing shoot, from one of the ordinary thallus cells. It curves backwards towards its point of origin and adjoins the basal cell from which the branch takes its rise. The tricho- gyne is also bent outwards, and penetrating the outer gelatinous coat of the thallus emerges into the open. Simultaneously with the formation of the carpogonial branch, the auxiliary cells are formed in the following manner. Two cells adjoin- ing the one which bears the carpogonial branch each segment off externally one cell in such fashion that the carpogonial branch lies between them. The parent cells of these two auxiliary cells are usually large thallus cells, but in no way distinct from the ordinary thallus cells. Sometimes both, sometimes one only is connected by a pore with the cell which bears the carpogonial branch. Though both of the auxiliary cells are apparently equally adapted for its function, only one of them is utilised, and it happens exceptionally that only one is actually produced. In most cases the fertilised carpogonium is directed towards the auxiliary cell destined to be used, and the latter is then rich . in contents. Immediately after fertilisation the carpogonium fuses first with the cell of its branch next it, and eventually with all four cells of the carpogonial branch, while at the same time all increase in size, as well as the cell which bears the branch. The great cell arising by this fusion is then farther increased by union with the cell that bears it, and only one large nucleus is discernible for RHODOPHYCE^, OR FLORIDE^ 227 the whole. While these changes are taking place the auxiliary cell also increases in size and richness of contents, while it approximates closely to the carpogonial cell. Sometimes, though not often, it C"*N ^\ .&£ /i f FIG. 72.-P/ocamuun corallorhiza natural size. emits a conjugation -process towards the carpogonial cell, and when this is present the nucleus is to be seen within the process. More frequently the carpo- gonial cell emits a process (ooblastema filament) towards the auxiliary cell. Ultimately fusion takes Q 2 228 SEAWEEDS place with the auxiliary cell and the nuclei of both unite. The formation of the wall of the cystocarp begins when the carpogonium has been fertilised and the auxiliary cell is recognisable. Branched filaments arise from the thallus-cells in the immediate vicinity of the one which bears the carpogonial branch, and their apices meet over the auxiliary cell and the whole body assumes a globular form. It is note- worthy that these filaments have no pore connections with the carpogonial or the auxiliary cell. They develop into the cystocarp wall, which at first so confines and presses against the swelling carpogonial and auxiliary cell that the contents of the wall-cells next adjoining die off and their membranes swell up. At about this stage the auxiliary cell, after union with the carpogonial cell, fuses with its parent cell, and the whole .united cylindrical body extends a kind of foot into the thallus beneath it. It may now for convenience sake be called the central cell. From its apical region there are now produced a number of marginal (gonimoblast) cells which have pore-connections with the central cell, and each of these gives rise to a carpospore. The central cell pro- ceeds to unite below with cells of the thallus adjoining it, while it bears more gonimoblast cells on its upper and middle portions. In the ripe fruit only the outermost layer of cells of the wall remains, and these swell up and yield at the apex, permitting the egress of the carpospores. Minor differences occur in the allied genera Cham- pia and Lomentaria. Among these it may be noted RHODOPHYCE.E, OR FLORIDE^ 229 that in L0wentaria clavellosa the fused carpogonial branch in uniting with the auxiliary cell first segments off a small cell which fuses with the auxiliary cell, but in L. articulata union takes place directly. Again, while in Chylodadia only single gonimoblast cells are given off by the central cell FIG. 73.—C)iampia parvula. a, b, c, d, successive stages in germination of carpospore ; e, optical section of further stage in segmentation of spore ; /, longitudinal section of apex of young plant highly magnified. (After Davis ) for the formation of carpospores, in Champia and Lomentaria multicellular branched gonimoblast fila- ments are given off, and the terminal cells of these give rise to carpospores. Mr. B. M. Davis has made a minute study of the development of the frond of Champia parvula from 230 SEAWEEDS the carpospore, the early stages of which may be seen represented in Fig. 73. The Geographical Distribution of the family is a wide one throughout both the north and south temperate zones and the tropical belt. Representa- tives penetrate also into the colder waters. Rhody- mcnia, Cordylecladia, Lomcntaria, Champia, Chyloda- dia and Plocamium are all British and for the most part abundant on our shores. Dclessericce. This family includes a number of the most beautiful red seaweeds, if indeed they are not the most beautiful of all Algae. They possess leaf-like fronds, some of them with midribs (Dclcsscria) others with delicate lace-like or net-like expansions (Claudea, Martensia, Vanvoorstia, genera formerly reckoned among the Rhodomelew) and are all notable for their conspicuous and graceful forms. Nitophyllum and its immediate allies (see p. 202) differ from most other Rhodophyccw in the occurrence of subsequent interca- lary divisions of the thallus filaments. The procarpia are situated in the middle layer of the fronds, and the gonimoblast produces the carpospores within a fruit cavity formed of the thickened cortical layer of the thallus, perforated in the centre for the escape of the spores. The gonimoblast is somewhat indistinctly divided into several lobes, formed simultaneously or in succession, which bear the carpospores singly or in series terminally, or more rarely almost all the cells form spores. FIG. 74. — Claudea elegant, a, frond natural size ; 6, part of frond with cystocarys very slightly enlarged. 232 SEAWEEDS The Geographical Distribution is well marked. Nitophyllum, Delesseria, and Hydrolapathum, the British genera, are widely represented, the two former in southern warm temperate seas as well as in the north. Claudea elegans occurs in Tasmania, and one small inconspicuous species, C. multifida, in Ceylon. Martensia occurs in South Africa, and Grinnellia is FIG. 75. — Martensia elegans reduced. confined to the Atlantic shores of North America. Delesseria Ainboinensis is a singular freshwater form occurring in running streams at a considerable elevation above sea-level in Amboyna. Bonnemaisoniecc. This small family includes for the most part forms with long slender main shoots clothed with fine filamentous branches. The procarpia are situated in RHODOPHYCE^E, OR FLORIDE.E 233 the cortical layers, and the fruit cavity is of similar character to that of the Delesscriecc. The gonimo- blast is a copiously branching tuft of filaments of which the terminal cells bear large club-shaped carpospores. Bonnemaisonia asparagoides is the only British form. Asparagopsis Delilei occurs in the warm North Atlantic ; and of the other noteworthy genera Ptilonia and Delisea occur in southern seas. Rhodomelece. This is one of the most natural and best defined families of Rhodophycece, not only from its repro- ductive characters but to a considerable extent its vegetative structure as well. The thallus in most cases consists of tiers of cells in series, a central one with smaller pericentral cells of the same length grouped round it. Branching, commonly of a mono- podial type, occurs, and the whole shoot is clothed more or less with fine hair-leaves if they may be so termed. Both antheridia and procarpia are formed on these hair-leaves, and are in the great majority of cases stalked. The cystocarps are therefore rarely sessile. In Polysiphonia, which may be taken as fairly typical of the rest of the family, the carpo- gonial branch is four or five-celled. The lowest of these cells becomes the auxiliary cell and the carpogonial branch so bends round that the car- pogonium itself touches the auxiliary cell (Fig. 77). The other joint-cells adjoining the one which gave origin to the carpogonial branch 234 SEAWEEDS divide and branch repeatedly, and so produce a body of cells enclosing the carpogonial branch. By division and growth of the cells that envelop this procarpium there is formed the wall . of the more or less globular fruit, perforated at the apex FIG. 76. — Cliftoncea pectinata natural size. FIG. 77. — a, Diagram of procarp in EhodomelefK (b is the axial cell, a is the cell from which the carpogonial branch eeec arises, and d is a branch cell which gives rise to sterile filaments); b, car- pogonial branch of Polysiphonia (after Schmitz) ; c, Dasya elegans reduced. and enclosing the gonimoblast. The gonimoblast is commonly a much suppressed tuft of cells on the upper, arched surface of which the carpospores are borne. These are generally ovate or club-shaped, RHODOPHYCE.E, OR FLORIDE^ 235 very rarely two or three small, round spores in series, and they escape through the apical opening of the stalked fruit. The tetraspores are very frequently formed in stichidia. The family is a large one consisting of a considerable number of genera of conspicuous size and beauty of form, and some of the genera, such as Polysiphonia, Z/aurencia, and Dasya, are rich in species. Pleuro&ichidium, from New Zealand, is however an exception, being a minute epiphyte of quite different habit from the other genera, but of essentially similar structure. The Geographical Distribution is world wide in the sea ; some of the genera, such as Polysiphonia, have a range as wide as that of the family. The British genera. Bostrychia, Rhodomcla, Odonthalia, Laurencia, Halopithys, Chondria, Polysiphonia,. Pterosiphonia, and Dasya, are for the most part very common on our shores. Ceramiece. This is one of the largest and most widely distri- buted families of Red Seaweeds, and members of it are everywhere common on coasts where Algal life is found at all. The thallus consists of single branched cell-filaments, sometimes with a cortex formed of rhizoid filaments (not a true cellular cor- tex). This false cortex is produced in the genera allied to Callithamnion by the outgrowth of rhizoid filaments from the basal cell of lateral branches, while in Ceramium and its immediate allies, such filaments spring from the upper ends (Fig. 78, «) of the cells of the thallus, forming a peripheral 236 SEAWEEDS crown of cells covering the place where the parent cell is joined with the one above it. In some cases these filaments completely clothe the main cell row, in others they fail to meet and leave intervening parts bare. The procarpia are therefore external, and the gonimoblast is usually destitute of any envelope, but such does occur. The gonimoblast (of which there are usually two in each cystocarp) forms successively several lobes, of which nearly all the cells give rise to spores. Callithamnion , one of the commonest genera, may be taken as typical of the others in the matter of its reproduction. From one of the joint cells of the thallus there issue (besides the vegetative lateral shoot from its upper end) two small unicellular outgrowths from about the middle of the cell. From one of these a three or four-celled carpogonial branch proceeds (not from the thallus-cell itself, as FIG. 78. — a, Ceramium diapha- ,, L i\ s-\ xi r ntim with nodal cells ; ft, procarp USUaliy Stated). Un the ier- of Callithamnion gracillimnm .,. . ~ •, highly magnified. (5, after tlllSatlOn of the CarpOgOU- Schmitz.) . ,-, -i i n / mm, the basal cells (or more rarely one of them) of the two original lateral outgrowths segment off each a daughter-cell, and these become the auxiliary cells. The fertilised carpogonium then extends a little sideways towards the auxiliary cells, and it also segments off its pro- truding parts as separate cells. These then conjugate with the auxiliary cells by means of an exceedingly OR 237 short ooblastema process. When the auxiliary cells have been thus fertilised they both divide, and the upper portions give rise to the two gonimoblasts of the cystocarp, while the lower portions fuse with the cells that bear them, and in some cases with the thallus-cell from which the fertile shoot originated. oob FIG. 79. — GloiosipJionin capillaris, fertilisation and development of cystocarp ; stages in order of numbers, a, auxiliary cell of procarpial branch ; ft, basal cell ; h, hypogonous cell of carpogonial branch ; 006, ooblastema — in (4) protoplast only ; sp, pollinoid ; t, trichogyne. With some slight modifications this process is valid for the numerous genera and very numerous species of Ceramicce. It includes such well-known genera, in addition to Callithamnion* as Lcjolisia, Fpermo- thamnion* SpJiondylothamnioii* Ptilothamnion* Griffithsia* ffalimis* Bornetia* Monospwa* Pleono- sporium* Haloplcgma, Ptilota* Ccmpsothamnion* 238 SEAWEEDS Plnmaria* Ballia, Antithamnion* Crouania* Spyridia* Carpoblcpharis, Ceramium* Microcladia* Rhodochorton* and Thamnocarpus, those with an asterisk occurring in British seas. The family is of universal distribution in the sea, especially the genera Callithamnion and Ceramium. CRYPTONEMIACE.E. General Characters. — As in the Gelidiece, the fertilised carpogonium emits a relatively long ooblastema filament, which branches copiously in the thallus tissue. Its terminal cells or joint-cells conjugate each with single auxiliary cells, and from these the gonimoblasts spring. In Gdidiem the fila- ment which emerges from the carpogonium is itself the goni- moblast, and its conjugations with thallus cells appear to be of nutritive importance only, while in the Cryptmicmiaccce the fused cells give rise to the gonimoblast. Glviosiphoni&z, Gratcloupiece, Dumontiecc, and Nemastomece. FIG. 80. — Young procarp „,, n P .,- •»-,• ^/^ of Gioiosiphonia capiiiaris. These four families, rigs. 71), 80, 81, are all characterised by the auxiliary cells being joint cells of secondary or primary filaments, and the carpogonial filaments usually of similar origin. The ooblastema filament PLATE VIII. 1. POLYIDES ROTUNDUS. 2. MELOBESIA MEMBRANACEA. 3. DUMONTIA FILIFOllMIS. 4. CORALLINA OFFICINALIS. Murray's Seaweed Plate VIII Hanharb imp RHODOPHYCE^E, OR FLORIDE^E 239 proceeding from the carpogonium successively ferti- lises a number of auxiliary cells, each produced on its own branch. After the conjugation of the ooblastema filaments and auxiliary cells, the gonimoblast that results is usually developed in the form of more or less definite gonimolobes, almost all the cells of which give rise to carpospores. The gonimoblast is in all cases within the thallus tissue. The geographical a FIG. 81 — a, fertilised carpogonium of Dudresnaya purpurifera ; b, ooblas- tema filament attached to auxiliary cell in passage (D. coccinea) highly mag- nified. (After Schinitz.) distribution is fairly general, and the British genera are Gloiosiphmiia (Gloiosiplwnicce) ; Halymcnia and Grateloupia (Grateloupiece) ; Dumontia, Dudresnaya, and Dilsca (Dumontiece) ; and Calosiphonia, Schizy- mcnia, and Furccllaria (Nemastomece). Rhizophyllidece. This family, also a small one of fairly general distribution, is represented, in British seas by only one genus, Polyidcs. The essential process of fertili- 240 SEAWEEDS sation is the same in this genus as in Dudrcsnaya (Fig. 81). The carpogonial branches and the cor- responding but more numerous branches that bear the auxiliary cells, are found together in special fertile portions of the cortex of the thallus. After fertilisation, this cortical tissue, having undergone FIG. 82.— Polyides rotundus. a, procarp with trichogyne and ooblastema ; b, ooblastema filament fertilising auxiliary cell highly magnified. (After Thuret and Bornet.) considerable development, contains close together numerous gonimoblasts, nearly all the cells of which eventually give rise to carpospores. Squamaricce. The thallus of the Squamariece is commonly minute and encrusting or consists of flat foliar expansions of tissue, in most cases encrusted with carbonate of lime. In this case, as in Polyidcs, the carpogonial branches RHODOPHYCE.E, OR FLORIDE^ 241 and the auxiliary cells (joint cells of ordinary thallus filaments) are associated together in fertile regions of the cortex. The gonimoblasts are very minute and numerous, and produce carpospores in nearly all their cells. The family, though a small one, is of general distribution, and is represented in British seas by the genera Petrocelis, Cruoria, Peyssoncllia,t Hcematocelis, and Rhododermis. Comllincce. Of all Rhodophycem this family is the most easily recognised, from the strong incrustation of the thallus with carbonate of lime, producing a stony consistency. The different forms are. however, of the most various outward appearance. As in Nitophyllum, so here the thallus filaments undergo subsequent intercalary divisions. Melobesia is of encrusting habit, like some of the Squamariece, growing by. marginal initial cells, at first circular, but afterwards becoming lobed and irregular through unequal development. Litho- phylhim forms thin stony plates of erect habit, while Lithothamnion gives rise to massive stony branches (Fig. 83). Starting from a stone or shell, which the thallus subsequently encloses more or less as a kind of core, its branches frequently form massive structures, in some cases almost rivalling the animal corals in bulk. They occur in particular abundance with the true corals, and the species of Lithothamnion, Melo- besia, &c., often act as a kind of mortar in holding together the reef-building corals. Corallina, Am- phiroa, Jania, and Cheilosporwn, are beautiful, stony, R 242 SEAWEEDS branched forms of a brittle character. The geological history of the family goes back to the Cretaceous FIG. 83. — Lithothamnion polymorphum reduced. times, and forms have been described, but with less certainty, from the early secondary rocks (Muschel- kalk, Trias). They first appear, however, beyond RHODOPHYCE^E, OR FLORIDE^E 243 doubt in the Senonian beds (near the top of the Cre- taceous beds) as Lithothamnion, and are thus of the same proved antiquity as the Diatoms. It does not occur massively until the Tertiary rocks are reached, Fir,. 84. — CoraW-nfi mediterrnnea, vertical section through antheridium highly magnified. (After Thuret and Bornet.) when this contemporary genus occurs abundantly in the lower Eocene. The Leitha limestone (Miocene), the Pisolite limestone, and the nummulitic rocks, are largely composed of Litkotkamnion. Though of greater abundance in warm tropical K 2 244 SEAWEEDS seas, the Corallines have a much wider range than the animal corals, and occur in considerable numbers in the colder regions of the ocean. On British shores, Corallina, Jania, Lithothamnion, LithopJiyllum, Melo- besia, and a very minute form, Schmitziella (incrusting Cladophora pellucida), represent the family with a fair number of species. The stony incrustation extends to all parts except the reproductive organs. Tetraspores, antheridia, and carpospores, are all formed in special conceptacular bodies (Figs. 84 and 85). In the flat encrusting forms these conceptacles appear as minute, wart-like outgrowths from the sterile part of the thallus ; in Corallina they occupy the summits of the branches ; and in Amphiroa they are lateral. In Lithothamnion these conceptacles are eventually overgrown by the increasing growth in thickness of the thallus, and in breaking down the stony substance they may be met with as small cavities representing the conceptacles of earlier periods of growth. The carpogonial branches occur, together with numerous auxiliary cells, in special fertile portions of the cortex. The auxiliary cells are joint cells of peculiarly differentiated thallus filaments. On the fertilisation of the carpogonium all, or nearly all, the auxiliary cells near it become fused with it by means of the ooblastema filament, forming a large ''conjuga- tion cell." From its periphery minute gonimoblasts arise, bearing chains of carpospores. The cystocarp is here, therefore, a kind of syncarp, since it results from the combination of numerous auxiliary cells and their products into one common definite fruit, RHODOPHYCE^, OR FLORIDE.E 245 FIG. 85. — Corallina mediterranea. a, section of conceptacle with tetra- spores ; b, ditto of cystocarp highly magnified. (After Thuret and Bornet.) 246 SEAWEEDS enclosed within a common cystocarp wall, with a single apical opening As has been said, the antheridia and tetraspores are produced within similar conceptacles. The tetrasporangia arise at the base of the cavity, and either surround a central sterile bundle of filaments, or they occur in groups, separated by a wall of tissue from each other, and each with a separate apical pore (Melobesict). BANGIACE.E. General Characters. — The thallus consists of cell- filaments (Bangia), or flat plates of one layer of cells and of irregular outline. Male reproductive bodies without cilia are produced within cells of the thallus, and they fertilise female cells specially distinguished from the thallus cells only by a short lateral protuberance. Non-sexual reproduction is effected by spores produced by the thallus cells, sometimes singly, sometimes after division of the parent cell. There has been so much discussion as to the systematic position of the Bangiacece, and with so uncertain a result, that the terminology of the repro- ductive bodies is difficult of application. The claim put forward for their inclusion among the Rhodo- phycece does not appear to be fully established, while at the same time there is no little difficulty in assigning them a place among the Chlorophycccu. The Thallus. — The main argument for the inclusion of the Bangiacecc among the Rhodophycecc is derived RHODOPHYCE^, OR FLORIDA 247 from the colour of the chromatophores, which is in perfect agreement, but this is of purely physiological significance. On the other hand intercalary divisions occur in the cells of the thallus of Bangiacccc, and though these occur also in Nitophyllum, Corallincce, &c., the argument is not in favour of inclusion on this ground. The pits between the thallus cells of Ehodophycccc are also absent from Bangiacccc. Beyond colour therefore there is little support to be gained from the character of the thallus in favour of inclusion. The Reproductive Organs. — The cells that become the female reproductive organs are indistinguishable from the ordinary thallus cells. The process of special- isation consists in a slight increase in size, but of no particular alteration in shape, except the acquisition of a very short lateral protuberance, called the homologue of a trichogyne by those who favour the inclusion of the order among the red seaweeds. The antheridia produce motionless unciliated pollinoids. In Ery- throtrichia one of the thallus cells produces a small superficial cell, which is segmented off and becomes the mother-cell of a pollinoid. There is here a certain resemblance to the Rlwdophycecc, but in Bangia and Porphyra the pollinoids are produced by the repeated division (in all directions) of one of the thallus cells which has gradually lost its colour. Numerous small pollinoids are thus produced and set free by the dis- solution of the membranes of the parent cells. The pollinoid fertilises the female cell on attaching itself to the short protuberance (trichogyne). An open communication is effected, and the contents of the 248 SEAWEEDS pollinoid pass into and unite with those of the female cell. There is no development of gonimoblast, but the whole of this fertilised cell becomes a spore and emerges from its membrane, or it first divides once or twice and forms several spores. The whole of this pro- cess may represent a very much reduced or an ancestral Rhodophycean type, but there is a great gulf between it and the simplest form of indubitable Red Seaweed. The non-sexual spores, the so-called tetraspores, of Bangiacece leave us also in doubt. The whole of the contents of a single thallus cell go to the formation of one of the spores which are unciliated, at first without a membrane, but afterwards with one. In some cases there is a preliminary division (once or twice) of the thallus cell. It may be recalled that monospores (the undivided tetrasporangium) occur in certain Rhodophycecc, but even then there is no conclusive evidence here for or against, though it leans towards inclusion. On the whole, and con- sidering the difficulty of placing them elsewhere, the Bangiacece may be left beside the Rhodcphycea, though not within the group. The Geographical Distribution is world wide. It is however a small order with comparatively few species, though these are of very variable character. Bangia, Goniot-richum, Erythrotrichia, Parphyra, and Diplo- derma, are all British, and most of them abundant on our shores. The species of Porphym furnish the edible Laver. SUB-CLASS V CYANOPHYCE^ THE primitive forms of Algae classified under this name possess in all cases a thallus of much sim- plicity, being unicellular, or composed of single rows of cells, nearly always embedded in definite gelatin- ous sheaths or gelatinous masses of indefinite out- ward form. The individual plants are in most cases associated together in colonies, the tendency to form gelatinous envelopes causing them to cohere in this fashion. Reproduction is typically a process of division of the thallus cells, though the precise mode of it, and of the liberation of the propagative bodies so formed, varies in the groups into which the Cyanophycccv are divided. In Chroococcaccce the cells are transformed into sporangia. A power of movement is exhibited, in the absence of cilia, by many members of the group, especially in the pro- pagative cells, but this power is sometimes retained by the mature thallus, as in the case of Oscillaria. The colouring matter of the cells is a bluish-green substance, phycocyanine, in addition to chlorophyll. 250 SEAWEEDS Other colours, such as purple, reddish purple, violet, yellow, and brown, are imparted to the plants by the coloration of their gelatinous envelopes. This colouring matter of the envelope has been called by Nsegeli scytoneminc, and is sometimes to be seen in the whole length of the sheath, sometimes only in part of it, but the peripheral part is more strongly tinged than the internal part, and it is displayed most vividly on the parts most exposed to light. The blue and red colours found in Homocystccc and Chroococcacece are absent from the Hcterocystece. The colouring matter of the cells is not associated with definite chromatophores, as has been reported by several observers, the error having arisen from the presence of crystalloid bodies, or from the fact of the plant studied not being a member of the Cyanophyc&B. The protoplasmic contents of the cells are almost uniformly tinged with the colouring matter. This colour differs somewhat with the age of the plant and the degree of its exposure to light. It is ordinarily more greenish in young fila- ments with uncoloured envelopes, but with age it becomes more olive or even yellow. As has been pointed out by Bornet and Flahault, the other colours, seen in the protoplasm of herbarium speci- mens, are due to decomposition. No true nucleus has been observed, though its discovery has been reported by several observers. Zacharias and others have shown that the error is due to the presence in the centre of the cells of a colourless portion of protoplasm, which may be stained with hsematoxy- lene. Its form, however, is not definitely limited CYANOPHYCE^E 251 like a true nucleus, and it has not been observed to display karyokinesis. Vacuoles do not appear ordinarily in the young cells, but with age, obscura- tion of light, or other unfavourable conditions, they arise and occupy a considerable part of the cell. It is not known, however, whether these vacuoles con- tain a cell-sap like those of other plants. Glycogen has been determined as present in the cells, but not starch. If we consider the close relationship of the Cyanophycecc to the Bacteria, it is not strange that theories of their polymorphism have arisen. It has been supposed for example that forms like those of Chroococcacccc are often stages in the development of the higher Cyanophyccce, and there are sufficient re- semblances to give colour to such a view. But there is no more proof of it than this slight ground of speculation, except the equally slender support de- rived from the fact that the forms frequently grow in the same places. It requires actual observation of development to establish such a matter. NOSTOCACE.E. This order is distinguished from the Chroococcacecc by its multicellular thallus, and by the production of hormogonia, formed by the fragmentation of the filaments into mobile segments. The whole of a cell row is called a trichome, and the trichome with its envelope, which may be gelatinous, or even almost cartilaginous in consistence, is called the filament. 252 SEAWEEDS It is most convenient to consider the Nostocaccce under two families, viz. the Heterocystecc and the Homocystcce. Heterocystcm. General Characters. — The cells of the trichome are differentiated into vegetative cells and into hctcrocysts, or cells incapable of farther development. The elongation of the trichome is by transverse division of all the cells (tribe Nostoceoe), or of a meristematic group of cells (tribe Eivulariece). There is true branching only in the tribe of Sirosiphonicce, effected by the division of cells parallel to the axis. Hormo- gonia, and spores endowed with the power of resting and of thus preserving the plant during unfavourable seasons and periods of dryness in particular, are the characteristic modes of reproduction. The Thallus. — The vegetative cells of the trichome vary least in the tribe Nostoccce, where they all appear to be very much alike ; differences appear in those forms classified under Scytonemccc, and are most marked of all in the tribe of Rimdaricce. The cells at the tip of a filament are generally shorter than at the base, where they attain an elongate cylindrical form. Cell-division takes place ordinarily when the cell attains its maximum length, but when growth is active the divisions succeed each other before that is reached and while the cells are still short. The envelope of the trichome (called the sheath in all the Nostocacecv) may be mucilaginous, gelatinous, or car- tilaginous in consistence, and there is considerable variety in its form and other characters, of service in CYANOPHYCE^ 253 classifying the genera and species, the thicker kind being frequently lamellated. The sheaths are at first colourless and transparent, and may remain so, but more often they become coloured as described. The FIG. 80.— a, Bivularia hospita ; b and c, hormogonia of Calotlirix pulvinata highly magnified. (After Bornet and Thuret.) heterocysts are special cells situated at the base of the trichome or intercalated in its course. They are bright green or light yellow in colour, have very little solid contents, and are commonly much larger than 254 SEAWEEDS the ordinary vegetative cells, from which they may be farther distinguished by their greater refraction. They adhere to the sheath, and at the point of their attachment to the neighbouring cells there is a little button-like projection of cell- wall. Besides the true branching in the tribe of Sirosiphoniece there are false branchings in other forms, occasioned by modes of growth, by the sticking of the heterocysts to the sheath, and by the development of hormogonia which have not escaped from the sheath, or have become fixed on the filaments. In the last case, of course, there is no sort of regularity. Reproduction is most frequently — is ordinarily — effected -by the production of hormogonia. These are mere segments of the trichome, to be distinguished from it with difficulty in some genera (e.g. Nostoc. Anabaina), but in others more specialised. In some cases the production of hormogonia terminates the existence of the thallus, and in fact involves its de- struction ; in others it begins at an early stage of development, while the filaments are still small, and proceeds more or less actively during the life of the plant. The hormogonia escape by sliding towards the opening at the end of the sheath. Some fix them- selves on the parent plant, but most descend to the substratum. On coming to rest they either develop at once into a new filament, or rest for a longer or shorter time while their cells increase in size and their sheath grows in thickness, sometimes becoming larger than the ordinary filaments. In other cases the hormogonia divide and subdivide, while they elongate and multiply for a time before assuming CYANOPHYCE.E 255 the characteristic form of the filaments of the parent plant. Spores are known in certain genera, and probably occur in all. They differ from the ordinary cells in their rounded form, greater size, and more granular contents, and their thick, coloured membrane. On germination the contents shrink from the wall and divide by parallel walls into a number of cells, and this short filament escapes by a perforation or cir- cumscission from the parent membrane. It then becomes practically a hormogonium, and thus repro- duces the plant. In a fresh-water form, Sacconcma, the spore divides by perpendicular walls and gives rise to a globular colony like Glo&ocapsa, but Borzi, who records the observation, did not observe the develop ment of this body into the normal form. Homocystece or Oscillariccc. General Characters. — The characters which dis- tinguish the Hmnocystece from the Heterocystem are mainly negative ; they have no heterocysts and no spores. Their reproduction is solely by hormogonia. The Thallus. — True branching, as in the Sirosi- phonc-ce, does not occur among the Hamocystecc, and since there are no heterocysts, there is no false branching of the type produced in the Hcterocystcce by the adhesion of the heterocyst to the sheath. Accordingly, in some of the genera the filaments remain simple, but false branching occurs frequently in the tribe Vaginariecc and in Plectoncma (Lyngbycce). The false branching in the former case arises when 256 SEAWEEDS several trichomes, or bundles of trichomes, diverge at the extremity of their common sheath, in which they remain partially embedded, while the free parts become clothed each with a separate sheath. Where two trichomes thus emerge, it sometimes occurs that a fairly regular dichotomous false branching is pro- duced. In the Lynglyece false branching (in spite of the absence of heterocysts), arises from a breaking of the trichome, sometimes by its mere length, some- times by a curvature of the filament, and the new ends breaking through the sheath grow out, or sometimes only one does this. The tribes Vagin- ariece and Lyngbyece have thus the first a terminal, the second a lateral branching. The trichomes grow at all points, but generally towards the apex there are indications of apical growth in a greater number of short cells. At the apex itself the terminal cell, more or less conical or like a cupola in shape, has a thick protective outer membrane, and its presence or absence and its precise form afford a systematic character in classification. The structure of the sheath has been very care- fully studied in this group by M. Gomont, and his observations are probably generally applicable to all Nostocaeca?; The sheath and cell-wall proper exhibit different chemical reactions, and while the latter appears to resemble cutine, the sheath consists of a substance nearly allied to cellulose. At the same time the sheath appears to become cutinised, when it becomes coloured under the influence of light. CYANOPHYCEJE 257 Since the Reproduction by hormogonia has already been described under the Heterocystecc, there is no distinctive character under this heading. The ab- sence of spores has also been noted above. The Geographical Distribution of Nostocaccm is world- wide in fresh-waters and damp situations, on coasts and in mid-ocean. They are of rare occurrence in cold situations, and the littoral marine forms are generally to be found near high-water mark. The calcareous forms are more characteristic of fresh- waters than of the sea, and are especially abundant in hot springs. When they live in waters strongly charged with lime, the precipitation of carbonate of lime is caused by the absorption of carbonic acid. However, this precipitation occurs only where the Alga3 do not find carbonic acid in sufficient quantity to meet the needs of vegetation. When it is abundant, as in certain mineral waters, the salt is precipitated only in very small quantities. The deposition may be round the individual filaments or so abundantly distributed as to envelop the whole thallus in a calcareous concretion. The genera Masti- gocolcus, Plcctcncma, and Phormidium, contain species which perforate shells and other calcareous bodies in the sea. The pelagic Oscillariem occur in enormous abundance in the warm waters of tropical seas. From the periodical occurrence of Trichodcsmium cry- thrcvum in great banks the Red Sea has obtained its name, and the same species and others allied to it have often been recorded from tropical seas in extraordin- ary floating masses. Most of the genera of Nostocacece which occur in the sea are known in British seas. 258 SEAWEEDS CHROOCOCCACE.E. General Characters. — The Chroococcacecc are dis- tinguished from the Nostocaccce not only by their unicellular character, but more particularly by the fact that they do not produce hormogonia, but uni- cellular reproductive cells. Typically cell-division does not occur in one direction only leading to the formation of trichomes, but the direction varies with more or less regularity, and since the daughter-cells remain together in colonies within the original envelope for a number of generations (Grl&ocapsa) irregular gelatinous masses are thus formed. The order is divided into two families, Chroococcccu and CUamcesiphonecc, which may for convenience be treated separately. The Chroococcece are best known by the genus Glceocapsa. After each division of the mother-cell the daughter-cells may be free to develop indepen- dently, but usually they are held together by the common gelatinous envelope for a succession of generations. Spores are formed after the simul- taneous change in habit of the whole colony of cells, and in place of the gelatinous membrane a thick membrane, rough on the outside, is developed. The spore repeatedly divides, and soon forms in this way a new normal colony of vegetative cells. The Chamcesiphonccc possess in Hyclla a genus of perforating Algae, which at first sight appears to be an approach to the Nostocacece, since the cells occur in filaments. They are, however, so many individual FIG 87. — a, Dermocarpa Schou*b"> Coccospheres (Fig. 58B), 185, 186, 189 Codiacece, 132 Codiolum, 178, 179 Codiophylliim, 144 Corftum (Plate III.) (Fig. 35), 132, 134, 135, 136, 137, 138 Coilodesme, 108 Colacohpis, 220 Collecting, 27 Colours, 4 Compsothamnion, 237 Conceptacle, 48, 73 CoraUina (Plate VIII.) (Figs. 84 and 85), 241, 244 CoralKmif, 241 Corallopsis, 224 Cordyltdadia, 230 Costaria, 76, 79 Craticular state, 190 Crouania, 237 Cruoria, 241 Cryptes piliferes, 53 Cryptonemiacew, 238 Cryptostoma, 52 Ctenocladns, 173 Culture of seaweeds, 8 Currents as agents of distribution, 10 Citfferia (Plate I. ) (Fig. 8), 56, 57, 58, 59, 60, 65, 68, 69, 117 INDEX 267 Cutleriacew, 50 Cyanophycece, 249 Oymathofre, 70, 79, 85 Cymopolia, 151, 152, 155 Cyatodonium (Plate VI.), 221 Cystophora, 55 CystophyUum, 55 Cystowxra (Plate I. ), 54, 55 Dactylopora, 154 Dactyl 'oporelt 'a, 154 Z>a.s^a (Fig. 77), 235 Dasydadacew, 145 JJasycladecc, 151 Daxydadns, 151, 152, 154, 155 DecaixneHa, 154 Decalcifying, 30 Delesseria (Plate VII.), 230, 232 Delesserieff, 230 Delisea, 233 Depth, range in, 5 Derbevia, 131, 132, 133, 134, 130, 137 Dermocarpa (Fig. 87), 260, 261 Detmarestia (Fig. 24), 99, 101 De$marestiacea>, 99 De*motrichum, 108 Diatomacew, 188 Diatomine, 189 Dictyoneuron, 77, 78, 79, 85 Dictyopteris, 60, 61, 63 Dictyoxiphon, 98, 99 J)ictyosiphonace, 252 Hildbrandtia, 200 HimanthaJia (Fig. 4), 45, 48, 49, 54, 55 ffomocystew, 255 Hormogonia, 251 fformoKira, 45, 47, 55 Hydrodathrus (Fig. 27), 54, 105, 108 Hydrolapathum, 232 /fyeffa (Fig. 88), 258, 261 Hypnea, 224 77ca, 175, 176 Isthmopha, 119 Jama, 241, 244 Kjellmania (Fig. 25), 102, 104 Laminaria (Fig. 13), 79, 85 Laminariacetv, 75 Landsbnryhia, 55 Laurencia, 235 Leathesia (Fig. 21), 91, 93 Lejolisia, 237 Lemanea, 200, 208 Leptonema, 95 Lessonia, 76, 77, 79, 81, 85 Letterstedtia (Fig. 54), 174, 176 Liagora, 209, 210 Light, interception of by sea- water, 5 Literature, 34 Lithoderma, 109, 110, 111 Lithophyllum, 241, 244 Lifhothamnion (Fig. 83), 241, 243, 244 Litosiphon, 108 Lomentaria (Plate VII.), 224, 225, 228, 229, 230 Lyngbye.ce, 255 Macrocystis (Figs. 16 and 17), 76, 77, 79, 81, 85 Marginaria, 55 Martensia (Fig. 75), 230, 232 Mastigocoleus, 257, 260 Melobesia (Plate VIII.), 241, 244, 246 Mesoglcea, 93 Microdadia, 238 Microdidyon, 156, 163, 165 Microsponginm, 91 Monospora, 237 Monoatroma, 174, 176 Mounting-fluid, 30 Munieria, 154 Myelopkyc&Si 108 Myriad is, 93 Myriodctdia, 93 Myriodesma, 48, 55 Myrionema, 91, 92, 93 INDEX 269 Myriotrichia (Fig 23), 95, 117 Naccaria (Plate V.) (Fig. 65), 215 Nemcdion (Plate V.j, 209, 210 Nemalionacetv , 207 Nemastomecc, 238 Nematophycus, 143 Neomeris (Fig. 45), 152, 155 Nereia, 86, 87 Nereocystis, 77, 79, 85 Nitophyllum, 202, 230, 232 Node, 190 Nodule, 190 Notheia (Fig. 5), 42, 47, 49, 51 Nostoc, 254 Noakocaccvt, 251 Nostocece, 252 Odonthalia, 235 Ooblastema filament (Fig. 82), 204 Oogonium (Figs. 2 and 34), 49, 58 61, 128 Oscillaria, 249 Oscillariece, 255 (Plate II.), 61, 63 Palmophyllum, 180, 181 Pelagic Algae, 17 Pelvetia (Plate L), 51, 55, 68 Penicillus (Fig. 38), 137, 138, 139 140, 144, 161 Peridiniece, 181 Perithalia, 86, 87 Perizonium, 194 Petrocdis, 241 Petrospongium, 92, 93 Peyssonellia, 241 Phacelomrpus (Fig. 70), 224 PhaopMa, 170, 172, 173 Ph, 39 Phhvospora, 102, 104 Phloiocaulon, 115 Phormidium, 257 Phycocelis, 117 Phycocyanine, 4 Phycoerythrine, 4 Phycophaeine, 4 Phycoxanthine, 4 PhyllUie, 108 Phyllophora (Plate VI.), 216, 219, 220 Phyllospora, 47, 55 Pinnularia (Fig. 59) Planktoniella (Fig. 60) Plectonema, 255, 257 Pleonosporium, 237 Pleurocapsa, 179 Pleurococcacea1, 179 Pleurodadia, 115, 116 Pleurostichidium, 235 Plocamitim (Fig. 72), 224, 230 Plumaria, 237 Pollinoid (Fig. 62A), 63, 202 Polycystis, 261 Polyides (Plate VIII.) (Fig. 82), 239, 240 Polyphysa, 145, 147 Polydphonia, (Plate VII.) (Fig. 77), 233, 235 Polytrypa, 154 Porphyra (Plate V. ), 247, 248 Porphyroglossum, 215 Postelsia (Figs. 15 and 17), 77, 79, 81, 85 Prasiola, 180, 181 Preservation in spirits, 30 Pringsheimia, 174, 176 Protococcacece, 177 Pterodadia, 215 Pterosiphonia, 235 Pterygophora, 78, 85 Ptilonia, 233 Pti/ophora, 215 Ptilopogon, 115 Ptilota, 237 Ptilothamnion, 237 Punctaria, 105, 108 Pycnophycus, 55 Pylaiella, 119 Pyrocystis (Fig. 57), 184, 185 , 109, 110 108 Range in depth, 5 Raphe, 190 270 SEAWEEDS Reproduction, 25 Rhabdoliths, 186, 187 Jfhabdonema, 195 Rhdbdonia, 221 Rhabdospheres (Fig. 58A), 185, 186, 189 Rhipilia, 137 Rhipocephaliis (Fig. 39), 141 Rhizodonium, 166, 169 Rhizophyllidea, 239 Rhodochorton, 238 Rhododermis, 241 Rhodomela, 235 Rhodomdew, 233 Rhodophycece, 200 RhodophyUideoe, 220 Rhodophyllis, 221 Rhodymenia (Plate VII.), 224, 230 Rhodymeniacece, 222 Rhodymenieo', 224 Rividaria, (Fig. 86) Rivulariew, 252 Saccorhiza (Fig. 12), 54, 76, 81, 84, 85 Salinity, influence of, 9 Sarcop'hycus (Fig. 6), 48, 49, 51, 55, 59 fiargassum, 43, 48, 55 Scaphospora (Fig. 10), 66, 67, 68 Schizymenia, 239 SchmitzieHa, 244 Scinaia (Plate V.) (Fig. 63), 210, 213 Scytonemine, 250 Scytosiphon, 84, 108 Scytothalia, 55 Scytothamnus, 54, 98, 99 Siphonocladus, 156, 159, 163, 165 fiirosiphoniew, 252 Solieria, 221 Sorocarpns, 117, 119 fipatogloMtim, 61 Spermatochnus (Fig. 19), 87 fipermothamnion, 237 Sphacelaria (Fig. 28), 113, 114, 115 Sphacelariacece, 111 Sphacdla, 115 Sphcerococcece, 223 Sphcerococcm, 224 Sphondy/othamnioii, 237 Splachnidiacea, 70 Splachnidium (Fig. 11), 70, 81, 84, 107 Spongodadia, 161, 163 fiporochnacew, 86 tiporochmts (Plate II.), 86, 87 Spyridia, 238 Squamariea, 240 Stenogramme (Fig. 66), 216, 217, 219 Stichidia, 206 Stictyosiphon (Fig. 26), 104 Stilophora (Fig. 20), 87, 89 StcKchospermum, 61 Streb/onema, 116, 119 Striaria, 102, 104 Striariacecc, 101 ^r^^^7ea (Fig. 48), 161, 163, 1C5 Stypocaulon, 113, 114 SuJma, 215 Suture, 190 Sykidion, 178, 179 Taonia, 60, 61, 63 Temperature, influence of, 7 Tetraspore (Fig. 61), 64, 205 Thalassiophyllum, 76, 79 Thamnocarpus, 238 Thysanodadia, 221 Tihpteridacea*, 66 Tifopteris, 66, 67, 68, 69 Tissues, 24 Trichodesm iinn, 257 Trichogyne (Fig. 62s), 203 Trichome, 251 TriploporeUa, 154 Tttomeya, 200 Turbinaria (Fig. 7), 44, 47, 48, 54, 55 (Fig. 40), 137, 138, 141, 143, 144 Udoteacecv, 137 INDEX Ulopteryx, 76, 78, 79 Ulothrix, 170, 172, 173 Ulotrichaceic, 170 Ulva (Plate IV.), 174, 176 Ulvacece, 174 Uteria, 154 Urospora (Fig. 51), 166, 168, 169, 170 Valve, 189 Vanvoorstia, 230 Vancheria (Fig. 34), 120, 127, 130, 132, 133, 136 Vaucheriacew, 127 Wrangelia, 215 Xiphophora, 55 Vaginariew, 255 Fa/onia (Fig. 46), 121, 156, 163, 164 ValoniacecK, 156 Zanardinia, 56, 57, 58, 59, 60 Zonaria, 60, 61, 63 Zoochlorella, 180, 181 Zoslerocarpus, 104 THE END RICHARD CLAY AXD SONS, HM1TEP, LONDON AND BUNGAY. Koyal Svo, bound iu half-roxburgh, gilt top, price £1 5s. net. PHYCOLOGICAL MEMOIRS: "Being researches made in the Botanical Department of the British Museum. EDITED BY GEOHGE MURRAY, F.R.S.E., F.L.S, With twenty Lithographic Plates. PARTS I., II. and III. London, 1892-95. DULAU AND CO., 37 Sono SQUARE, \v.