1891.] MICROSCOPICAL JOUENAL. 83 



Tetniemortis : 



granulatus, ... " 3S- 50 " 



Xanthidiiim : 



armatum, .... " 63-140 '• 



cristatum, . . . . " 40- 55 " 



Arthrodesinus : 



incus, . . . . " 10- 36 " 



octocornis, ... " 16- 35 " 



Euastnun : 



ansatum, . . . . . " 35- 36 '• 



elegans, .... " 18- 36 " 



binale, . . . . " 15- 25 " 



Micrasto'ias : 



radiosa, . . . . ' " 150-200 '' 



truncata, . . . . " 50-100 " 



Jenneri, . . . . '' 100-150 '• 



pinnatifida, ... " 83-110'" 



Staui'astrtim : 



pygniEeum, . . . " 16- 35 " 



pentacladum, . . . " 40- 70 *■' 



controversum, . . . " 38- 6<^ " 



Brasiliense, . . . " 87-130 " 



Eustephanum, . . . " 50- 75 " 



Comparing these figures the variations will be found to range from 

 1-3|. That is to say, the sa77ie species may include individuals ^ larger 

 than others, ^ larger, f larger, twice as large, yes, three times as large 

 in diameter. The desmids, then, do not seem to keep their size much 

 better than the diatoms in spite of the fact that they have no silic'eous 

 casing to contend with. In many cases it is true the variation may. not 

 be due to decrease, but however caused it is a departure from any fixed 

 size. I simply wish to point out that all individuals between these ex- 

 tremes are classed as belonging to the same species. Allowing, then, 

 the same amount of variation to the diatoms it is evident that we might 

 have a large number of frustules which would well represent the specific 

 size even after repeated subdivisions. As the size only goes down in 

 one of the pair of frustules at each division, and then only by a portion 

 of the diameter equal to double the thickness of the hoop, it will take 

 quite a number of divisions to use up the allowable extent of variation 

 from -3—31. For all those variations cited in the case of 40 desmids 

 are stated as occurring in individuals which are classed by Wolle as 

 species, not varieties. The diameter o{ Xanthidium armatum^ for ex- 

 ample, is 62-140 <). ; of Euastrum elegans^ 18-36 P- \ of Arthrodesmiis 

 inctts, 10-36 //.. If a desmid can vary that much, and yet all the vari- 

 ations be classed as members of the same species, not varieties, it is only 

 fair to allow the same range for a diatom . The difference in size therefore 

 of individuals of the same species among the diatoms is not sO remark- 

 able after all. If some are a third, a half larger than others, twice as 

 large, three times as large, the same is just as true of the desmids. 



But wait ! I believe I have made a discovery. It would seem as if 

 there must be something wrong again either about the received theory 

 or the drawings in many instances. For by the theory the diatoms must 



