Phylogeny of Desmidiacese 377 



The third hypothesis is that the Desmidiacese are a homogeneous group and that all 

 Desmids primarily arose from filamentous ancestors. There is little doubt that the great 

 majority of Desmids have arisen in this way, and since the various members of the 

 Spirotfeniese, which is the chief group of Saccoderm Desmids, are more closely allied to 

 the Placoderm Desmids than to the Zygnemacese there appears to be little reason for 

 supposing the family Desmidiacese (in its widest sense) to be a mixed assemblage. This 

 third hypothesis is the one supported by a considerable amount of evidence and it seems 

 on the whole to be the most probable (consult scheme B). 



There are many cogent reasons for regarding the Desmidiacese as a 

 specialized family of Conjugates which has originated from filamentous 

 ancestors. The loss of the filamentous condition has been accompanied by 

 the development of complex morphological characters, and this has gone on 

 hand in hand with the loss of sexual differentiation of the conjugating cells 

 (G. S. W., '99 ; '04). The structure of the cell-wall of the Placoderm Desmids, 

 which is so much more complex than anything exhibited in the Zygnemacese, 

 and the elaboration of the chloroplasts, clearly indicate that the Desmidiacess 

 are not primitive. 



It has been previously mentioned that Desmidium cylindricum is the only 

 known Desmid in which the zygospore is lodged in the female gametangium 

 (consult p. 372 and fig. 235 J) and the very occasional reversion to this type 

 of conjugation in Hyalotheca dissiliens (Joshua, '82 ; Boldt, '88) is of con- 

 siderable significance. It is in Desmidium cylindricum and in the presumed 

 'abnormal' cases of conjugation in Hyalotheca dissiliens that one is probably 

 witnessing the type of conjugation which was prevalent in the ancestors of 

 the Desmidiacere. Another Desmid of great interest is Debarya desmidioides 

 (W. & G. S. W., '03), since it stands in a somewhat intermediate position 

 between Mesottenium and such genera as Mougeotia and Debarya. 



The average Desmid must be regarded as a unit of a dismembered filament. 

 The complete individuality of the cell is the only real distinction between 

 Desmids and other Conjugates. Lutman ('11) from his observations on the 

 division of Closterium Ehrenbergii states that ' the position of the young 

 transverse wall would seem to indicate that the pointed ends are secondarily 

 formed, and that Closterium was originally a filamentous alga, which has 

 developed the habit of breaking up into single cells.' 



Desmids exhibit a considerable tendency towards the secondary assump- 

 tion of the filamentous condition. This has resulted in the production, 

 from unicells, of filamentous genera, such as Streptonema, Onychonema, 

 Sphcvrozosma, Desmidium, etc., in which special apical connecting processes 

 are for the most part conspicuously developed (vide fig. 224 A, J and N; and 

 fig. 225 A H). The same tendency sometimes occurs in certain species of 

 genera which are normally unicellular. Thus, short filaments of cells have 

 been figured in such species as Cosmarium obliguum Nordst. (73), C. monili- 

 forme (Turp.) Ralfs, G. Regnellii Wille (W. & G. S. W., '96), Euastrum binale 



