18 FOKMATlOISr OF SPOEES OF EHIZOFUS AND PHYCOMYCES. 



that there are A'ery manj- exceeding-ly small ones, with detinitely 

 rounded outlines. Most of these are scarcely larger than the nuclei, 

 and some are much smaller. They can not, therefore, be shown in a 

 drawing on so small a scale as PI. 1, tig. (>. They are, however, essen- 

 tially the same in size, number, and distribution as those shown in 

 PI. il, tig. 9. 



Thus far, except for the arrangement of the cytoplasm and nuclei, we 

 have had no phenomena in the sporangium that even suggest cell divi- 

 sion, unless possibly it be the clear zone. The greater part of the more 

 solid portion of the cytoplasm has formed itself into a layer at the 

 periphery. Nearly all of the nuclei also have migrated into this por- 

 tion of the sporangium, and are distributed irregularly throughout 

 the dense cytoplasm. They are not even approximately' equidistant 

 from each other, nor are the}' often, if ever, in actual contact, though 

 Leger states that such is very frequently the case. How he could 

 determine the normal distribution of the nuclei from crushed sporan- 

 gia is difficult to comprehend. 



As soon as the protoplasm is distributed as has been described, the 

 separation of that which is to be included within the columella from 

 that which is to form the spores begins. The columella is not at first 

 a flat cross wall at the base of the sporangium which is later pushed up 

 by turgor to its characteristic dome shape, as it is currently described 

 as doing, but is laid down in essentiallv the same fashion as described 

 by Harper (1899) for Piloholus. There first appears in the denser 

 plasm a single layer of spherical vacuoles (PI. H, fig. 7) running par- 

 allel to its inner surface. The layer of the denser plasm inside the 

 system of vacuoles is usually from one-fifteenth to one-twentieth as 

 thick as the layer outside. Apparently these vacuoles are formed by 

 the enlaro-ement Of the very minute ones already mentioned that lie 

 in this region, rather than by the migration of previously enlarged 

 vacuoles. In sporangia in which this layer of vacuoles is only partly 

 formed there are usually a few large vacuoles arranged in the layer, 

 and between them are smaller ones, varying in size down to the small- 

 est in the sporangium (PI. II, fig. 7). This leads one to believe that 

 the vacuoles in this layer are essentially like the others in the sporan- 

 gium and in the mycelium. These vacuoles and all others in the spo- 

 rangium agree with those of Piloholus and Sjwrodinia in being devoid 

 of all stainable contents (PI. II, fig. 7), in which respect they difier 

 strikingly from those of Phyeomyces^ described later. 



The vacuoles are at first spherical, or nearly so, but soon begin to 

 flatten, their long axes being parallel to the inner surface of the denser 

 plasm. Bv this flattening the}' become disk-shaped, as in PL II, fig. 8, 

 and the edges of adjacent ones come in contact and fuse, forming a 

 narrow curved cleft in the protoplasm. At the same time a circular 

 furrow begins to cut upward from the surface of the protoplasm at 



