168 DIVISION II.—COURSE OF DEVELOPMENT OF FUNGI. 
is gradually becoming finely granular, it divides into the numerous sporangia which 
form the sorus (c,d). The swarm-cells (g) discharged from the sporangia become new 
resting-cells, if they find a host. This is the history, for example, of Synchytrium 
Mercurialis, S. Anemones, &c. ‘The course of development in the Eusynchytrieae 
is somewhat different; an unlimited number of generations of sori may be and 
actually are interpolated between every two successive hybernating generations 
of resting-cells; these sori are formed from swarm-spores without there being 
any resting period and produce again new swarm-spores. The spore which has 
penetrated into the swelling cell of its host developes into the thin-walled initial cell, 
which divides into the members of the sorus as soon as it has reached a definite stage 
of development. After a number of such generations have been produced, partly 
perhaps owing to external causes, the formation of resting-cells recommences. The 
behaviour in germination of these also varies in different species; in some, as for 
instance Synchytrium Stellariae (Fig. 78) and S. Oenotherae, they form sori like 
those of the Pycnochytrieae; in others each resting-cell becomes a single sporangium 
directly without forming sori, as in S. Taraxaci. 

FIG. 78. Synchytrium Stellariac, Schrot., from Stellarıa media, a—d germination of the resting-cells when placed 
in water after hybernating in the dry state. a—c successive states of the same specimen, 4 four-hours later than @, c seven days 
later and 'five days after division into the cells of thesorus. da mature sorus with the mature sporangia falling asunder. e—g a 
single iumin water. ei diately after being taken from the sorus. J the same two hours later shortly before formation 
of spores, the small sphere of fatty substance being divided into parts of equal size. g forty-five minutes later with the swarm- 
cells escaping. a—d magn. 145, e—g 375 times. 

It is obvious that the difference between the Synchytrieae and the group of the 
Olpidieae lies in the formation of sori. Species like Synchytrium Taraxaci with 
resting-cells which become sporangia without dividing are to some extent intermediate 
forms. Neither conjugation nor any sexual process has been observed in either group. 
From the facts before us we can only gather that a sorus or resting-cell proceeds 
directly from a swarm-spore. Influenced by the analogy of the isogamous Algae and 
Protomyces, I sought carefully in Synchytrium Stellariae and S. aureum for the con- 
jugation of free swarm-cells, but, as was the case in my examination of Cladochytrium, 
I never found an instance of it, either between spores of one or of different sporangia’ 
belonging to the same or to different sori. The abnormally large swarm-cells which 
are observed occasionally with two or more spheres of fatty matter and additional 
cilia are monstrous forms and not products of conjugation. These negative results 
do not preclude the possibility of conjugation; it might occur inside the cell of the 
host if more than one swarm-cell had found its way there. Cornu is inclined to 
account in this way for the formation of the resting-cells of Synchytrium, but he 
adduces no decisive observations in support of his view, but only an arbitrary com- 
parison of states of S. Stellariae found side by side. The objections to this are 
so obvious that they need not be specified. Direct observation of the development 
