376 Conjugate 



As pointed out elsewhere (G. S. W., '15), there is no essential family difference between a 

 Mesotsenium and a Closterium. The fact that there is no obvious line of junction between 

 the new and the old semicells in the Spirotaenieae is of little importance, since the develop- 

 ment of the new semicells is the same as in Closterium (compare fig. 231 A and D with 

 fig. 231 F and /) and other Desmids. Moreover, Oltmanns' further contention that in 

 those Desmids which he would place in the ' Mesotaeniacese ' the germination of the zygo- 

 spore results in the production of four embryos is not entirely supported by facts. 



Only one true case of hybridization has been observed amongst Desmids and in that 

 case the development of the zygospore was not followed out (Archer, '75). In this case a 

 zygospore had been produced by the conjugation of two cells, one of which was Euastrum 

 Didelta (Turp.) Ralfs and the other Euastrum hum.erosum Ralfs. All other recorded cases 

 of hybrids are conjectural and most of them are obviously forms produced by ordinary 

 vegetative division. It is unlikely that many true hybrids can exist among the Des- 

 midiacese, since sexual reproduction in the group as a whole is exceedingly rare and in 

 very many forms is not known to occur. 



Concerning the phylogeny of Desmids and their position in the Conjugate 

 more than one view has been put forward. The present author is convinced 

 that they constitute a highly specialized family, degenerate in sexual charac- 

 ters, but with an amplification of morphological characters unsurpassed in 

 any other group of unicells. 



ZYGNEMACE.E 



ZYGNEMACEAE 





DESMIDIACE^E S 1 I \ 



Debarya desmidioides /rt , , -. ., . / / Dpharva 



, (Placoderm Desmids) / / **/' 



| / J desmidioides 



^tro^n^ / Gonatozyg* 



r includes certain Algae ~| ? ^acco^eym^g Placodermx 



Lremoved from Desmidiace* J DESMIDIACE^ 



Scheme A Scheme B 



Three possibilities may be discussed in considering the relationships between Desmids 

 and other Conjugates. The first is that the unicellular Desmids are primitive and that 

 the Zygnemaceae have been derived from them. This hypothesis may be dismissed, since 

 there is strong evidence that many Desmids have originated from filamentous ancestors 

 and no definite evidence that any Desmid has arisen from a unicellular ancestor. 



The second possibility is that the Desmidiaceae are not a natural family and that the 

 Spirotseniese, and even the Gonatozygae, should be removed from them. It might then be 

 suggested that the Spirotaenieae ( = the Mesotaeniacese of Oltmanns) arose from unicellular 

 ancestors and in course of time gave rise to such forms as Debarya desmidioides, through 

 which and similar forms the Zygnemaceae were evolved. The great bulk of the Desmidiaceae 

 (all the Placoderm Desmids) would then be derived from the Zygnemaceae by retrogression 

 and specialization (consult scheme A). 



