I'llYI.DI.I N 1 



101 



anterior in position. Secondly, schizogony of the 

 Plasmodium has not been reported in "'. polycystic, 



anil nothing is known about the types of nuclear 



divisions in the vegetative and sporogeneous thalli. 

 Schiaogony and the occurrence of "promitosis" are 

 claimed to be outstanding characteristics of the 

 Plasmodiophorales. Thirdly, the sporangia and rest 

 i nii spores of W. polycystic give a positive cellulose 

 reaction when tested with chloro-iodide of sine, 



while those of tile Plasmodiophorales do not. Fur- 

 thermore, in germination the content of the zoospore 

 • liters the host directly through a penetration tube, 



leaving the empty spore ease on the outside of the 



host cell as in Olpidiopsis, RoBellopsis and other 

 similar genera. In the Plasmodiophorales. on the 

 other hand, the zoospores are reported to enter the 

 host directly as a naked amoeboid body. How sig- 

 nificant these minor differences are in phylogeny and 

 whether or not they outweigh the similarities in 

 thallus structure and type of development remains 

 to be seen from future studies. Nevertheless, the 

 presence or absence of cellulose is regarded as fun- 

 damentally significant by many students of phylog- 

 eny and evolution. Jf'oronina glomerata differs 

 from the previous species in several ways. Instead 

 of forming hut one zoospore in germination as in 

 W. polycystic and the plasniodiophoraceous species, 

 the resting spore functions as a sporangium and pro- 

 duces a large number of zoospores. Furthermore, the 

 Plasmodium is animal-like in mode of nutrition. 

 according to Zopf ('91) and Scherffel ('25). and 

 engulfs plastids. starch grains, and other solid 

 bodies. This material is digested in well-defined food 

 vacuoles, and the extraneous waste material is dis- 

 carded to the outside in preparation for sporogene- 

 sis. This type of feeding and digestion is character- 

 istic of the Proteomyxa. and for this reason Zopf 

 and Scherffel relegated W, glomerata to the zoo- 

 spore group of the Myxozoidia or Proteomyxa. 

 Thus, within the same genus, as Woronina is now 

 interpreted, occur species with seemingly diverse 

 relationships. However, these differences of rela- 

 tionship may not prove to he as significant as they 

 now appear, because evidence is accumulating which 



suggests that certain species of tin- Proteomyxa, 



Plasmodiophorales. and Woronina may possibly be 



closely related. 



Turning to the other two genera. Pyrrhosorus 

 and Rozellopsis, which are temporarily included in 

 tin- Woroninaceae. it becomes evident that the rela- 

 tionships are not well defined. Zoosporangia and 

 resting spores apparently do not occur in Pyrrho- 

 SOmS. Instead, the plasiuodium cleaves into spore 

 mother cells which unite into a sorus and later 

 undergo three divisions, forming eight free spores. 

 The- latter are transformed directly into birlagellate- 

 isocont zoospores. Despite these- differences the 

 presence of a plasmodium and sorus suggests some 

 degree of relation to or parallelism in development 



with the Plasmodiophorales. .1 ml I '01) was uncer- 

 tain of the relationship of Pyrrhosorus, but he em- 

 phasized the striking similarity of its method of 



spore development to that of Tetramyxa. As tin 



writer ("1-2) has already pointed out. had Octo 

 myxa been known at that time, duel would doubt 

 less have emphasized the relationship of his fungus 



with the Plasmodiophorales even more strongly, 



Winge ('IS) also regarded it as clo8ely related to 

 the- Plasmodiophoraceae and made extensive com- 

 parisons between its life cycle and that of Sorolpi 

 (Hum. He considered the- sporangiosori of the latter 

 genus as homologous with the sori of spore- mother 



cells of Pyrrhosorus and believed that the absence 



of walls around the spore mother cells is of minor 



significance. Cook ('33). on the other hand, be 



lieved that the relation of this genus to the Plasmo 

 diophorales is very questionable. 



The origin and relationships of the provisional 

 genus Rozellopsis are even more obscure. Compact 



or loose sporangio- and cystosori are unknown, and 

 the only significant characters which it has in com- 

 mon with the two previous genera are its plasmo- 

 dium-like vegetative thallus. which may or may not 

 undergo schizogony or division, and birlagellate 

 zoospores. Since- the presence of a plasmodium has 

 not been conclusively demonstrated in this genus, 

 the inclusion of Rosellopsis in the same family with 

 Jl oronina and Pyrrhosorus becomes even more ques- 

 tionable. However, the anteriorly biflagellate hetero 

 cont zoospores of R. simulans, according to Tokuna- 

 ga's ('33) drawings, are strikingly similar to those 

 of the Plasmodiophorales, but in infecting the host 

 they behave like those of II'. polycystis, Olpidiopsis, 

 Ectrogella, etc. Instead of entering the host direct 

 lv. they form an infection tube through which their 

 content passes into the host cells. On the other hand, 

 the structure and development of the thallus and 

 resting spores in the aseptigenous and septigenous 

 species are identical to those of the mono- and poly- 

 sporangiate species, respectively, of the chytrid 

 genus. Rosella and Pringsheimella dioica, as far as 

 is now known. Whether this indicates merely a 

 parallelism in development from different ancestors 

 or direct relationship is not certain. According to 

 Bessey's ('42) theory of origin through the reten- 

 tion or loss of the- second flagelluin. Rozellopsis is 

 more primitive than Rozella and Pringsheimella and 

 may have given rise to these genera by the loss of 

 one flagellum. While this theory se-ems plausible, it 

 is obvious that the loss of one flagellum, without 

 change in position of the remaining one. from the 



zoospore of Rozellopsis would not lead directly to 

 the distinctly posteriorly uniflagellate zoospore of 

 Rozella and Pringsheimella, since- both flagella in 



Rozellopsis are reported to be either late-ral e>r 

 anterior in position. Loss of one flagellum in R. siinu- 

 lans, for instance-, would make the zoospore ante- 

 riorly uniflagellate-. 



The- other four families to be- considered, namely, 



the- Kctrogellaceae. Olpidiopsidaceae, Sirolpidia 



ceae, and Lagenidiaceae, appear t < > be more- closely 



related as a whole- and constitute- an aseiniling or 

 descending line-, depending on which viewpoint one 

 holds. The- principal genera of these- families e-on 



