The Classification of the Algae. 85

much stretch of imagination be considered reconstructed stem-forms, which inhabited Greater New Zealand or elsewhere duringthe earlier Pliocene or earlier still. As to whether, the normalspiny form of the Discaria ana the artificial spineless form, i.e. theprolonged seedless-form, are really two distinct species, or merelythe two extremes of one variable species, opens up far too wide aquestion for consideration here. It is obvious however, that thepresence of either or both in a region would be entirely a matterof climate, station and competition with other plants and animals.

In conclusion, I must express my thanks to Professor CharlesChilton, D Sc. for his kmdness in allowing me to use the moistchamber of the Biological School, Canterbury College, Christ-church, New Zealand, for the completion of the above mentionedexperiment.

Wellington, New Zealand,> « . 23rd, 1905.

DESCRIPTION OF THE FIGURES ON PLATE II. ILLUSTRATING

DR. L. COCKAYNE'S PAPER ON THE SIGNIFICANCE OF SPINES

IN DISCARIA TOMATOU.

Fig 1. —Photograph of seedling plant of Discaria after twelve months' culturein moist chamber. On left is adult shoot placed on pot for purposesof comparison with seedling.

Fig. 2.—Photograph of Discaria in characteristic habitat. Fan of Creek fromMr. Torlesse, Eastern climatic region of Canterbury, S, Island of N.Z.Altitude about 700m. (Both photographs by the Author).

THE CLASSIFICATION OF THE ALGAE.

A REVIEW OF PROFESSOR OLTHANNS' RECENT BOOK.'

AFTER the large amount of work of all kinds that has beenpublished on the Algae, more particularly on the green forms

and on the plant-like Flagellata, during the last fifteen years, aputting together of the results in one or more general works hasbecome urgently necessary to enable the student to appreciate the

' Morphologie und Biologie der Algen, von Dr. FriedrichOltmanns. Erster Band, Spezieller Theil. Gustav. Fischer,Jena, 1904. Pp. VI. and 733, with 476 figures in the text.

86 Review.

advances that have been made and to obtain a general view of thewhole body of our present knowledge.

Professor Oltmanns' bulky volume is the first (special) partonly of his projected " Handbuch," of which he hopes to publish thesecond (general) part in the course of this spring. The present partof this extensive work will in itself, however, be of very great useto Algologists. Its full and clear treatment, with excellent andabundant illustration, of the very large amount of data now available,render it by far the most notable general work of reference on thisfascinating group. At the same time it is full of interesting andsuggestive discussions. As the author remarks in his preface, nogeneral work has appeared since Falkenberg's Algae was publishedin Schenck's " Handbuch der Botanik " (1882), and though the partsof Engler and Prantl's " Natiirliche Pflanzenfamilien " dealing withthe Algae by various writers, have helped to bridge this long interval,yet the form of treatment necessarily imposed upon contributors tothat invaluable undertaking has prevented the work of Wille,Kjellman and Hauptfleisch on the green, brown and red algaerespectively, from quite filling the place of a detailed handbook ortreatise in which the author is free to devote any amount of spacehe may think desirable to general descriptions, comparisons anddiscussions. Furthermore our knowledge of the Algae hasincreased so enormously, even since the publication of the parts ofthe " Pflanzenfamilien " dealing with them, necessitating in somecases a revision of our fundamental conceptions of their morphology,that a fresh treatment is urgently required.

A significant feature of the change which has occurred in therelative importance to be attached to the various groups in con-sidering the morphology of the whole is the large amount of spacedevoted at the outset to the plant-like Flagellata, and to theunicellular forms generally. " The Algae," says Professor Oltmannsin his opening sentence, "go back phylogenetically to the Flagellata.That is a conviction which is continually making more headway . . . "And he goes on to point out that a recognition of this truth enablesone to give a fixed position to many forms which have for decadesbeen tossed backwards and forwards between Zoologists and Botanists.The work of Butschli, and particularly of Klebs, established thisfundamental starting-point for a consideration of the morphologyof the Algae beyond all doubt. Wille took up the main phylogeneticthread at the point where the algal characters become definitelypredominant, and developed its ramifications with considerable

The Classification of the Algae. 87

success. Later, the Swedish algologists, Luther and Bohlin, havegiven good reason for supposing that the Flagellata gave rise totrue algal descendants, not along one line only, but along severaland finally Senn's treatment of the Flagellata in the " Pflanzen-familien," though he does not accept Bohlin's conclusions, hasfurther served to emphasise the importance of this group asprecursors of the Algae.

Professor Oltmanns' treatment of the Flagellata is based onthat of Klebs and Senn. That curious "algal Flagellate" Hydrurus,he includes in the ChrysomonadineaE (the yellow-brown division ofthe Flagellata) and such forms as Phceocysiis, Pltceococcus and evenPhceothamnion (which in certain respects is quite a highly developedAlga and was actually placed by Wille among the pure greenfilamentous forms in the Chroolepideae) are treated as an appendixto the yellow-brown Flagellates. This is indeed the least objectionableposition for these lowly-organized fixed brown forms at present,since their cell-structure is in most cases typically flagellate. Thereis much to be said for considering them as efforts in the direction ofalgal evolution from a stock of Flagellates which ultimately gaverise to the Phaeophyceae proper, (Brown seaweeds), especially asthe motile cell in these simple forms often shows the typicalone-sided (monosymmetrical) structure characteristic of thezoospores and gametes of the Phaeophyceae; and this view isstrengthened by the great importance which we must now attach,as a result mainly of the studies of the Swedish algologists to thecharacters of the algal motile cell. For the time being we areinclined to agree with Professor Oltmanns' view (p. 13) that it istoo early to place these simple forms actually with the Phaeophycese,and they are too miscellaneous a collection to be given a family oftheir own.

The series of Heterokontas (spelt by our author with a c) ofBohlin is, we are glad to see, completely accepted by ProfessorOltmanns. We regard the building up of this series out of manywell-known and some newly discovered forms, a building up whichwas the cumulative work of Borzi, Lagerheim, Luther and finallyBohlin, as quite the most brilliant and notable advance in themorphology and classiflcation of the Green Algae that has been madefor many years. The striking cytological characters in which theforms assigned to this series agree :—their yellow-green pigment, con-tained usually in several discoid chromatophores without pyrenoids,the formation of oil instead of starch as an assimilative product, and

88 Review.

finally the characteristic motile form with two unequal flagella—serveat once to convince us of the natural character of the series and toextend the generalisation that constancy in cytological character isone of the most important of morphological features and thereforeone of the most important marks of affinity in the Algae, and shouldentirely overrule similarity of conformation and habit of the thallus,and even resemblance in the form of reproduction. This general-isation was already well established in regard to the pure green,unicellular and colonial forms, while among the Heterokontae wehave a series of organisms ranging from Chloramwba, a naked,flagellate, amoeboid form, through unicellular types which shew anincreasing preponderance of algal features, to coenocytic forms likeBotrydium, and multicellular filamentous forms like Conferva,(recently found to posses motile anisogametes). A striking peculiarityof the Heterokont^e, as opposed to the pure green series, is theapparent rarity of gamogenesis, which has not reached anything likeso high a stage of evolution as in the latter. Professor Oltmannsdoes not follow Bohlin in transferring the Vaucheriaceae to theHeterokontan series, and his decision is perhaps scarcely to bewondered at. Nevertheless the traditional position of Vauchcriaamong the Siphonales is certainly not satisfactory. With theexception of the formal resemblance given by the fact of its thallusconsisting of a branched ccenocytic tube, a resemblance on whichwe can lay scarcely any emphasis in the absence of other evidenceVaucheria scarcely presents a single character in common with theother Siphonales. Pyrenoids and starch are absent, and themethod of reproduction, both in the highly differentiated sexualprocess, and the curious and unique synzoosporcs, has no parallelwhatever in the other families. Ernst's recently describedDichotomosiphon, which forms starch and has traces of internalring-walls at the bases of the branches, but which in respect ofits reproduction is undoubtedly a Vaucheria, does certainly take ussome little way towards bridging the gulf, a gulf that neverthelessremains sufficiently wide. On the other hand we have a resemblanceto the HeterokontsE in the absence of starch and presence of an oilin Vaucheria itself, while the antherozoids of the latter, withtheir often unequal, and laterally inserted flagella, seem to work invery well with the Heterokontan type, though the cilia of theVaucherian synzoospore apparently fail to concur. On the wholewe incline to a belief in the Heterokontan alliance, thoughDichotomosiphon slightly weakens the case for such an affinity. Our

The Classification of the Algae. 89

author's account of the many existing observations on the structureandbehaviourofthiscuriousgroup is excellent and interesting, thoughwe note that Davis (Bot. Gaz. Vol. xxxviii.. No. 2,1904) has contradictedhis account of oogenesis, affirming that of the numerous nucleipresent in the yonng oogonium all but one are destroyed insteadof wandering back into the parent tube. Davis' version certainlyseems a priori more likely to be correct, but we ought to waitfor confirmation of one or other of the accounts before we canregard the matter as settled.

Perhaps the most important consequence flowing from theestablishment of the series of Heterokontas is the weight which itcompels us to attribute to the characters of the motile cell as a stablemorphological feature. Wille had already separated the Conjugataefrom the rest of the Green Algae on very good grounds, and with thefurther break-up of the remaining families according to theirzoospore-characters, we find the whole of the green forms (excludingthe Characese) falling into three great series, separated by numerousdistinctions of which the most striking is the locomotor apparatusof the reproductive cells. The only outstanding exceptions are theclosely allied genera CEdogoniuin and Bulboclicete, (with which mustbe associated Stahl's striking form (Edocladinin, described in 1891but apparently never seen since), and Derbesia, a siphoneous formusually placed close to Bryopsis on account of the strong resemblanceof its vegetative structure to that of the latter genus. The zoosporesof Derbesia, however, which were described by Solier in 1847 and ofwhich no more recent account is available, are said to resemble thoseof the CEdogoniacece in possessing a crown of very numerous ciliaattached round the anterior end, and therefore differ very strikinglyindeed from the bi-flagellate type, which (with quadri-flagellatevariations) is universal among the zoospores of the remaining greenforms. This fact led Bohlin to separate the CEdogoniaceae as aspecial series the Stephanokontae and to be logical, Derbesia shouldgo with them. Thus we have four great series of Green Algae,Isokontae with two equal flagella, including the great majority of thegenera, Heterokontte with two unequal flagella, Stephanokontae witha crown of cilia, and AkontEe with non-ciliated reproductive cells.The hypothesis is that each of these series or phyla—at any ratethe first three—is separately derived from the Flagellata, theancestors of each having the characteristic ciliation. In the caseof the first two we have actual evidence of such derivation.

The strength of the case for this view, which has already been

90 Review.briefly referred to, is increased by the fact (alluded to by Bohlin)that there is apparently no evidence of zoospores having increasedthe number of their cilia in the course of descent. There are infact actually no transitional forms between the types of ciliationdescribed, unless a certain amount of variation in length of theshorter flagellum of the Heterokontan type (in which the two flagellaare sometimes of almost equal length) and the frequent occurrenceof four flagella in the Isokontse can be so considered.

It is further increased by the parallel—though less strong—evidence of a similar independent derivation from the Flagellata ofthe Brown and the Red Algas. The possible connexion of the formerwith the Chrysomonadineae through forms like Phoiothamnion andPhceocystis has already been alluded to, and the similarity of theirmotile cells pointed out. There is absolutely no evidence of anyconnexion of the primitive members of the Phaeophyceas proper, e.g.the Ectocarpaceae, with green filamentous forms.

In the case of the red forms the evidence of the connexion ofthe Rhodophycese with the Flagellates is at present very slight, butprobably a better case could be made out for such a connexion thanhas yet been done, and we should certainly not be surprised at thediscovery of further evidence pointing in this direction.

The net result of all these considerations is the conviction thatthe cytological characters of the primitive motile cell (its chroma-tophore, pigment and product of assimilation—handed on of courseto the immotile cells of the thallus which have taken over thevegetative functions in the higher forms) and its locomotor apparatus,are of the first importance as constant morphological features andtherefore as characters to which the greatest taxonomic weightmust be attributed. We had hoped that Professor Oltmannswould have laid more stress than he has done on these points in soimportant a work as the one before us and we trust that he willgive us a full discussion of the general bearings of these topics inhis " Allgemeiner Theil."

After a useful summary of those groups of plant-like Flagellates,which are not clearly associated with algal forms—the Crypto-monadineae, the Euglenaceae, and the Dinoflagellata (with rather afull notice of Schutt's work on the cell-wall, etc. of the last named)Professor Oltmanns proceeds to treat the Conjugatae and the Diatomstogether under the name Acontae, a term introduced in the" Revision of the Classification of the Green Algae " published in thisjournal in 1902, and reprinted separately in 1903. This is anunfortunate proceeding. The term Akontae was intended to

The Classification of the Algae. 91

emphasise the phylogenetic importance of the motile cell andits consequences in accordance with Bohlin's principles to whichwe have just called attention afresh. In the view of the authors ofthe " Revision," the Conjugatae are an isolated group of pure greenforms whose origin is very doubtful and which are fundamentallydistinguished from the other " pure green " phyla by the fact thattheir reproductive cells have no ciliation. The possibility of theirunion with the Diatoms appears to the authors of the " Revision "to be excluded by the difference of pigment (apart from othercharacters) since all the evidence appears to point to pigment as ofabsolutely the first importance as a taxonomic character. SinceProfessor Oltmanns does not adhere to this view it would have beenbetter if he had avoided the term Akontae altogether and called hiscomposite Conjugatae-Bacillariales group simply Zygophyceae (aname he has given as an alternative). The use of Akontae in thissense introduces a further confusion in nomenclature which it wasthe object of the authors of the " Revision " to simplify on thepromising lines established by the Swedish Algologists.

In his classification of the difficult group of the Conjugatae,Professor Oltmanns deviates somewhat from previous arrangements.He includes Genicularia and Gonatozygon, usually placed with theDesmids, in the filamentous group (Zygnemaceae). This is a changein the same direction as that made in the " Revision " where thesegenera are placed in a separate family Archidesmidiaceae, intendedto connect the filamentous forms with the more specialised Desmids.Professor Ottmanns goes further, and himself makes a separategroup—the Mesotaeniaceae—in which he places the generaMesotiieniuin, Spirotaenia and Cylindrocystis, characterised by simplemembranes and the production of four embryos from each zygote.These three genera, though nearer the Desmids proper thanGenicularia and Gonatozygon, are certainly the least specialised formsof the latter. They exhibit the three types of chromatophore foundin the filamentous forms, while most of the other Desmids shewsome combination of plates and ridges. Professor Oltmannsconsiders that his Mesotaeniaceae are the simplest and mostprimitive Conjugates (p. 53) and (if Genieularia and Gonatozygonare placed with the filamentous forms) they certainly seem toconnect the Desmids and Zygnemales. He lays more stress on themode of conjugation than on the form of the chromatophore,derivingthe method of gamete union found in Closteriuiit from that describedby Archer in Spirotcenia and the Zygnemaceous type from thatfound in Cylindroeystis. He points out that a common character of

92 Review.

the Desmids proper is the production of two embryos from thezygote, while the Mesotaeniaceae have four and the Zygnemaceaeone.

In the filamentous forms Professor Oltmanns makes Debaryathe most primitive type on account of its mode of conjugation, andfrom this derives Zygnema, Spirogyra and Sirogoniuiu on the oneside and on the other Zygogonium and Mougeotia. He deliberatelyneglects the form of the chromatophore (which Palla made thebasis of a classification of the filamentous Conjugatae adoptedin the " Revision") and quite frankly refuses to discuss thesubject. This is, we think to be regretted, since the relativeimportance of such characters among the various groups of Algaeappears to be a most interesting and important topic. We cannotenter into it in detail here, but may say that while ProfessorOltmanns' derivation of the various modes of conjugation is bothinteresting and ingenious, on his own shewing these processessometimes vary to such an extent within the limits of a singlegenus or even of a single species that the propriety of using themto the exclusion of characters which are singularly constant withinthe genera must be seriously called in question ; while the absenceof any attempt to shew how one well-marked type of chromato-phore can arise from another, how for instance the Spirogvra-typecan be derived from the Debarya-type, is a serious defect whenprobable lines of evolution are being sketched out. We cannotresist the conviction that it is the cell-characters which have beenconstant for the longest time while the details of the conjugationof gametes, even the details of the behaviour of protoplasm uponwhich Professor Oltmanns lays stress, are much more likely to beeasily modified. We should all admit, of course, that the incipientdifl'erentiation of sex seen in Spirogyra is an advance on the per-fectly isogamous conjugation of Debarya or Mougeotia. The pointis that more taxonomic weight should be attached to differences ofcell-structure, because these appear to be more constant characters.

Having devoted so much space to a consideration of some ofthe more fundamental topics connected particularly with the moreprimitive green forms, there is but little left in which to do justiceto the rest of Professor Oltmanns' work.

We should much like, for instance, to discuss in detail hisclassiflcation of the unicellular "Isokontae." This certainly shewsa great advance on any previous classification put forward in ageneral work, though we are not in agreement with some of thedetails. However, as our author truly remarks, the making of a

The Classification of the Algae. 93new system of these forms has become a kind of sport to the Algo-logist, and while probably the great majority of existing genericforms have now been described, we are in need of a good deal moredetailed information as to their minute structure and life-historybefore we shall reach any sort of finality in their classification.They are certainly a most difficult group to deal with, since thegenera are so numerous, their external form so various, and theircell-structure often difficult to observe with sufficient accuracy,unless good cytological methods, difficult to apply in many cases,are employed. Professor Oltmanns, it is interesting to note,separates Volvocales (the motile) from Protococcales (the motion-less forms) and includes Tetrasporaceae (in a somewhat narrowsense) in the former. While there can be no doubt that some of themucilaginous immotile types are extremely closely related to theChlamydomonadines, it seems unnatural to separate them altogetherfrom numerous other unicellular immotile forms which our authorputs in Protococcales. Surely it is most natural to draw the lineat the point where the immotile phase of the life-history becomesdominant. Professor Oltmanns artificially simplifies his problem tosome extent by omitting all reference to a certain number of genera.The " Scenedesmaceze" practically correspond with the Selenas-traceae plus the Phytheliaceje of the " Revision," and are no doubta natural group. Of the remaining families there is not very muchthat need be said.

A group Ulotrichales is adopted by Professor Oltmanns toinclude the branched and unbranched filamentous types (theUlotrichales of the " Revision " together with the Ulvaceae and theCEdogoniaceas). Neither the parenchymatous structure of theformer, nor the characteristic motile cells and other peculiarities ofthe latter, seem to Professor Oltmanns to constitute sufficientreason for placing them in separate groups.

The genera included in the Chietophoraceae, whose limits differwidely from those given by Wille, mainly follow the arrangementadopted in the " Revision," in which, largely on the ground ofHuber's excellent researches, the epiphytic and endophytic formswere regarded as progressive reduction-series from the primitiveStigeocloniuin-type. In the present work however Aphanochcete isgiven a separate family on the ground of its sexual reproduction.It is interesting to note also that he is inclined to accept Chodat'sreport of the existence of aplanospores, zoospores and gametes inPleuroeoccus and to follow this author in considering the genus areduced Chaetophoraceous type. Certainly the cell-division and

94 Review.thread-formation and branching which occur in Plenroeoccusestablish some case for taking it out of the Protococcales, butwe can scarcely agree with a similar view of Gloeocystis. TheColeochaetaceae are rightly considered the highest segment of theChaetophoraceous series. The name Chroolepidaceae is confined tothe very natural little group of aerial forms—Trentepohlia, Phyca-peltis and Cephaleuros (the Chroolepideze of the " Revision.")

Professor Oltmanns divides the filamentous ccenocytic puregreen genera into two great groups : (1) Siphonocladiales includingCladophoraceae, {with Anadyomene, Microdictyon and Dictyosphaeria),Siphonocladiacese {Sipltonocladus, Chaiiicedoris and Struvea), Valo-niaceas [Vnlonia alone) and Dasycladaceae (with the usual genera),and (2) Siphonales. Though we entirely agree that the generacontained in the first three of these families should all be closelyassociated, the actual arrangement is unusual and seems insuf-ficiently justified in the text. The Dasycladaceae, though oftentechnically " siphonocladous " certainly stand apart from the othergroupinvirtue of their many peculiar characters. It is more probablethat they are derived from some strictly siphoneous verticillateform in the neighbourhood of Bryopsis than from any of the trueSiphonocladeae.

The treatment of Siphonales calls for no special remark.We must pass very lightly over the great groups of the Brown

and Red Seaweeds, though they actually occupy a little more thanhalf the work. The Phaeophyceae are divided into three primarydivisions—the Phaeosporaceae, Akinetosporeae and Cyclosporese.The Phaeosporeae are again divided into four great families—theEctocarpaceae, Cutleriaceae, Sphacelariaceae and Laminariaceae, andthen again into sub-families. This arrangement permits a muchreadier general view of this difficult group than is the case whenthe non-cyclosporous Phseophycese are arranged in a great numberof coordinated families. The Akinetosporeae are perhaps asomewhat provisional group, established to include the Tilopteridaceaeand Choristocarpaceae with motionless spores. In the Cyclosporeaewe have, of course, the two families, Dictyotaceae and Fucaceae.The discoveries of Lloyd Williams have certainly brought thesetwo families so much nearer together that it is no longer advisableto place them in separate cohorts as Engler does. Williams' mostrecent papers, however, with his striking discovery of a truealternation of generations in the Dictyotacese, were, we supposepublished too late to find a place in the present account. Thedescription of the Fucaceae is excellent, but we could have wished

The Classification of the Algae, 95

that the cytological details of reduction, etc., in oogenesis, couldhave been included. It is, we think, a mistake to divorce suchimportant details from the special descriptions of the groups theyrefer to, however necessary it may be to consider them again in thecourse of general discussions.

The Bangiales are treated as a separate group distinct fromthe Rhodophyceae, which name is taken as synonymous withFlorideae. The great feature of the treatment of the Red Sea-weeds is a section of no less than 110 pages devoted to adescription of their vegetative structure under two great heads—the " Springbrunnen-typus " and the " Zentralfaden-typus." Thisdetailed and yet generalised treatment enables one to obtain an" Uebersicht" of the structure of these forms in a way that hasnever been possible heretofore.

The reproductive processes of the Red Seaweeds, finally,occupy the concluding section of the work. Here ProfessorOltmanns is dealing with a subject, to which his own work hascontributed by far the most important advance of the last fifteenyears. We need only note that in his arrangement of the familiesbased on the structure of the sporophyte he follows Schmitz andHauptfieisch in the main, departing from Schmitz's groupingonly in comparatively minor points.

One of the principal risks to which an author is exposed inundertaking a full treatment of a group containing a very greatnumber of varied forms is the danger of becoming lost in details.We do not for a moment suggest that Professor Oltmanns hasactually fallen into this danger, but we do rather miss, particularlyin some parts of his book, that broad evolutionary treatment, whichwe think might have illuminated and given more unity to this solidand well-informed work. The principles governing the evolution ofplants are not always easy to disentangle, but in the case of theAlgae, particularly the lower green forms, which are not, like thehigher plants, burdened with a legacy of complex structure, thereis rather an exceptional opportunity of obtaining a clear viewof at least the proximate factors of evolution. It may be urgedthat this is scarcely a fair criticism to make on the " special part"of a work of which the " general part " is still unpublished, but thebroad treatment referred to is wanted in actual connexion with thedetails to give them life and unity. It is perhaps hypercritical,however, to receive so solid, useful and attractive a work in sucha spirit. We look forward to the author's " Allgemeiner Theil" withgreat interest. A.G.T.