12 



THE SIMPLE HOLOCARPIC BIFLAGELLATE PHYCOMYCETES 



pseudopod-like extensions and vacuoles (figs. 6, 7) 

 it is not certain from Juel's account that they move 

 about and migrate from cell to cell as in Plasmodio- 

 phora, etc. No evidence of schizogony was observed 

 by Juel, but Winge interpreted some of the uninu- 

 cleate stages as probable meronts. 



The mature plasmodium is multinucleate, vacuo- 

 late, and usually irregular in shape (figs. 6-8), and 

 just before sporulating forms an enveloping mem- 

 brane like Sorolpidium. Plasmodia which are exten- 

 sively drawn out and occupy several host cells may 

 accordingly appear lobed, irregular, and tubular 

 (fig. 12) after the wall has formed. Following this 

 stage the protoplasm divides into uninucleate seg- 

 ments. In this process no distinct cleavage furrows 

 have been observed. The plasmodium appears to be- 

 come highly vacuolate (fig. 8) during this process, 

 and the cytoplasm accumulates around the nuclei 

 and forms stellate protoplasmic islands which re- 

 semble somewhat the sporonts of Tetramyxa. These 

 segments soon become almost spherical or spindle- 

 shaped (fig. 12), and Juel thought that the latter 

 type of cells are formed in plasmodia which are 

 highly vacuolate and scarce in cytoplasm. In addi- 

 tion to these two kinds of segments, irregular elon- 

 gate, oval and smaller ones may be formed, appar- 

 ently as the result of unequal cleavage, which finally 

 degenerate. 



The spherical, 8 /x in diameter, and spindle- 

 shaped segments are uninucleate, naked, and never 

 develop a distinct wall. They aggregate to form a 

 definite sorus (fig. 9) and each cell soon divides into 

 octads of spores as in Octomyxa, which led Juel to 

 call them spore-mother cells. In this process of spore 

 formation the nuclei divide mitotically (figs. 15-18) 

 and each mitosis is followed by cell division. Defi- 

 nite chromosomes (2 to 5) are formed on a sharply- 

 defined spindle during mitosis, and there is no evi- 

 dence of "promitosis," according to Juel's figures. 

 Each of the eight naked spores soon becomes trans- 

 formed directly into zoospores without developing 

 thick walls and becoming dormant. The zoospores 

 apparently infect the host cells and develop into the 

 small thalli shown in figures 2 and 3. 



ROZELLOPSIS 



Karling, 1942. Aracr. Jour. Bot. 29: 33. My- 

 cologia 34 : 205. 



which extend through the host wall; usually filling 

 the host sporangia or the hypertrophied portions of 

 the hyphae completely; sporangium wall tightly 

 pressed against, seemingly fused with, and usually 

 indistinguishable from that of the host. Zoospores 

 slightly variable in size and shape, with one to sev- 

 eral minute globules, heterocont, shorter flagellum 

 usually extending forward and the longer one back- 

 ward in swimming; zoospores swirling in the spo- 

 rangium before emerging and swimming away ; con- 

 tent of zoospore flowing into host cell through an in- 

 fection tube in germination, leaving the empty zoo- 

 spore case attached to host cell. Resting spores un- 

 known in monosporangiate species ; solitary in septi- 

 genous polysporangiate species, lying free within 

 host cell and separate from host wall, variable in 

 size, brown and spiny; protoplasm coarsely granu- 

 lar, including a large vacuole or globule of hyaline 

 material; germination unknown. 



This genus was created for the Rozella-like spe- 

 cies with biflagellate heterocont zoospores which 

 have been described from time to time. As such it is 

 perhaps scarcely more than a provisional dumping 

 ground for imperfectly known species of this type. 

 It was proposed primarily to include Pleolpidium 

 inflatum Butler ('07) and a similar parasite which 

 Miss Waterhouse ('40) found in Phytophthora. 

 Whether or not the species which Fischer described 

 as R. septigena and R. simulant belong here is ob- 

 viously open to question. He figured and described 

 the zoospores as biflagellate and heterocont, but it 

 is particularly noted in this connection that his de- 

 scription does not apply specifically to these spe- 

 cies. It relates instead to the zoospores of Woronina, 

 Olpidiopsis and Rozella collectively. Inasmuch as 

 many of Fischer's observations of other similar para- 

 sites have proven inaccurate, it is not altogether im- 

 probable that he may have been mistaken about the 

 number, relative lengths, and position of the flagella. 

 On the other hand, it is equally probable that he had 

 at hand a different fungus from the one described by 

 Cornu as R. septigena. This is suggested by To- 

 kunaga's ('33) confirmation of Fischer's report of 

 biflagellate heterocont zoospores in R. simulant, 

 which is identical to R. septigena except in host 

 range. For this reason Fischer's R. septigena has 

 been separated from Cornu's species of the same 

 name and placed temporarily with R. simulans in 

 Rozellopsis. This genus accordingly includes two in- 

 completely known aseptigenous monosporangiate, 

 and two doubtful septigenous, polysporangiate spe- 



( PLATE 1) 



Thallus, intramatrical, more or less indistinguish- 

 able from but apparently immiscible with the host 

 protoplasm; becoming invested witli a wall at ma- 

 turity and forming one sporangium ; or cleaving (?) 

 into several segments which become separated by 

 host walls, mature in basipetal succession, and de- 

 velop into sporangia or resting spores. Sporangia 

 terminal or intercalary in host hyphae, variable in 

 size and shape, with one to several exit papillae 



cies. 



So far as is now known Rozcllopsis has the same 

 type of development as Rozella. In monosporangiate 

 species the thallus develops into one sporangium or 

 resting spore, whereas in the septigenous members 

 the thallus is reported to segment into several por- 

 tions, each of which develops into a sporangium or a 

 spore. Germination of the zoospores, infection, and 

 entrance of the parasite have not been observed in 

 R. inflata, so that the following description of the 

 processes is based on R. uaterhouseii and R. septi- 



