WITH NOTES ON OTHER SPECIES, 71 



the sporangia of Apliaiiomyces often reach a length of more than a hundred times 

 their diameter. 



In rare cases, the quantity of protoplasm contained in the sporangium may be 

 sufficient to completely fill it, but usually it forms a parietal layer of greater or less 

 thickness, with a vacuolar space extending through the middle. If this layer be very 

 thin, or if the sporangium be completely filled, that condition will induce certain 

 modifications in the usual course of development of the zoospores, but in a great 

 majority of cases the process is as follows. In consequence of the greater turgidity 

 of the sporangium than of the lower part of its hypha, its basal wall becomes convex 

 towards the base of the filament (Fig. 7, a). The first indications of the formation of 

 zoospores then soon follow. The phenomena attending this process have been the 

 subject of much study and of widely different interpretations, most of which cannot 

 profitably be detailed here. The most important contributions to the discussion 

 have been those by Strasburger ('80), Biisgen ('82), Ward ('83), Berthold ('8(5), 

 Hartog ('87), and Rothert ('88) ; and their papers may be consulted for the details of 

 the various views put forth. Repeated studies of several species have satisfied me 

 that Eothert's account, which is corroborated in most details by Berthold and 

 Ilartog, is practically correct. Therefore the following account is a combination 

 of the descriptions given by those writers with personal observations. The descrip- 

 tion may best be based, as has been said, on the commonest form of sporangium, 

 that with a parietal layer of protoplasm of considerable thickness and an axial 

 vacuole. At first, irregular rifts begin to appear in the protoplasm, extending out- 

 ward from the vacuole. They soon become more definite and more numerous, and 

 connect with each other in such fashion that the protoplasm is marked off into a 

 number of irregularly polygonal masses, as seen from the surface (Fig. 7, a). It is 

 probable that the number of these blocks, which finally become spores, corresponds 

 to the number of nuclei originally shut in by the basal wall, since the zoospores are 

 always uninucleate, and there is no evidence that any nuclear division occurs within 

 the sporangium. The clefts are at first quite narrow, and the protoplasmic masses, 

 or " spore origins," as they have been called, are frequently connected by threads of 

 protoplasm. The somewhat irregular outlines and the granular structure of the 

 origins, together with the appearance of the connecting threads in surface view, have 

 led Strasburger, Biisgen and Ward to very different interpretations of these clefts 

 from those here adopted. They have regarded them as " cell-plates," separating the 

 spores, and consisting of layers — lines in optical section — of granules. But it is 



A. p. S. — VOL. XVII. J. 



