82 



THE AMERICAN MONTHLY 



[April, 



some of the frustules which, in spite of repeated subdivision, would 

 keep their normal size, one for every individual of the species repre- 

 sented at the time division begins. And that of the size represented by 

 the smaller valve of each of those frustules there would be twice as 

 many. In other words, that only one of each succeeding pair of frus- 

 tules is smaller than the frustule from which the two are produced. 

 Moreover, that one will be smaller all around simply by the thickness 

 of the hoop at each division. Thus there will be a considerable num- 

 ber of frustules which will not vary much from the size of the diatom 

 when multiplication began. And I also said the result of subdivision 

 was often a decrease in size in the case of thedesmids and of the infusoria ; 

 that is, that such decrease was not confined to the diatoms, not merely the 

 result of a siliceous covering. On reading over that statement it occurred 

 to me to look up the extent of variation in individuals of the same 

 species among the desmids and I was surprised to find how wide it 

 was in many instances. Here are the figures for some 40 species cov- 

 ering 17 out of the 19 genera enumerated by Wolle : 



Hyalotheca : 



disilliens, . 



dubia, . 

 Bambusina : 



Brebissonii, 

 Desmidium : 



aptogonium, 

 Sphoerozosma : 



pulchrum, . 



serratum, 

 Mesotcenium : 



micrococcum, 

 Spirotcenia : 



obscura, 

 Peniuiti : 



digitus, 



lamellosum, 

 Closterium : 



obtusum, 



augustatum, 



striolatum, . 



Leibleinii, . 

 Docidium : 



crenulatum, 



truncatum, 



coronatum, 

 Calocylindrus : 



connatus, . 



minor, 

 Cosmarium : 



ovale, 



quadratum, 



pyramidatum, 



Ralfsii, 



Diam. 20- 36 // 

 " 13- 21 " 



i8- 25 - 



25- 38 - 



