222 



THE AMERICAN MONTHLY 



[December, 



angial frustules of the diatoms are 

 notoriously more coarsely marked 

 than the parent frustules. There are 

 a great many species of diatoms, be- 

 longing to the N. pinna group, which 

 really pass into each other so gra- 

 dually, that even by the help of stria- 

 tion it is difficult to distinguish them ; 

 N. affinis produces, by conjugation, 

 true N. pinna, and I have even 

 observed the large frustules of the 

 latter again producing monsters, by 

 conjugation, far more coarsely 

 marked than the parent frustules ; 

 shall we consider the sporangial form 

 as one species, and the parent form 

 another? 



I have before me now, a slide of 

 Gomphonema olivaceum containing 

 myriads of frustules, many conjugat- 

 ing, and some with the parent frustules 

 yet adhering to the sporangium. The 

 comparative striation, as measured 

 with a Powell & Leland spider-line 

 micrometer, is very nearly as 4 to 6, 

 and as the individual measurements 

 of the parent frustules give for the 

 striation 28 to 30 in .ooi-in.,we have 

 for the sporangial ones say about 20 in 

 .001 in. In this gathering there are 

 numerous free sporangial frustules, 

 wholly formed, and quite as coarsely 

 marked, and, apparently, numerous 

 others of intermediate size and stria- 

 tion — of what value would striation 

 be here ? What I have said about G. 

 olivaceum, is equally true of other dia- 

 toms, notably of the genus Cymbella. 

 And yet, in connection with other 

 characters the striation should not be 

 ignored. In the same gathering, on 

 Isthmia enervis, the striation may be 

 so nearly the same on larger and 

 smaller frustules as to appear to 

 be of specific value, but it by no 

 means follows that it will be the same 

 in this species from a widely dif- 

 ferent locality, nor does my exper- 

 ience with Eulenstein's preparations 

 of Isthmia enervis, coincide with that 

 of Count Castracane. I find that the 

 small granules on the connecting zone, 

 or central portion, say in .001-inch, in 

 the ratio of about 5 to 7, measuring, 



however, not with extreme accuracy, 

 yet sufficiently accurately to show 

 quite a latitude in this respect. Taking 

 a pretty pure gathering, made at the 

 time of the year somewhat remote 

 from the time of the conjugation, I am 

 quite prepared to admit that a prep- 

 aration of the so-called Frustulia Sax- 

 onica, for example, will not show any 

 appreciable difterence in the striation 

 of the frustules ; but I would be quite 

 unwilling to admit that this diatom 

 could not be obtained from another 

 locality, considerably more finely or 

 more coarsely marked; indeed, Count 

 Castracane himself admits a differ- 

 ence, though he says it has never, to 

 his knowledge, exceeded ^, which, as 

 Mr. Kitton shows, gives a range in 

 JV. crassinervis, if he understands 

 aright, of 27 to 35 in .001 an inch ! 



The general character of.the stria- 

 tion, parallel, radiate, etc., thecharac- 

 ter of the median line, if present, the 

 comparative fineness or coarseness of 

 the striae, — all these are, no doubt, 

 important, as is also within limited 

 range, the number of striae in .001 of 

 an inch. Anyone looking over Mr. 

 Habirshaw's " Catalogue of the Dia- 

 tomaceae " will realize what a fright- 

 ful increase of species was made by 

 Ehrenberg, and the earlier observers, 

 from considering the number of rays, 

 in the genus Actinocyclus as of 

 specific value, equally pernicious is 

 the custom too largely indulged in at 

 the present day by many hard-work- 

 ing Continental observers, who, look- 

 ing from the stand-point which Count 

 Castracane appears to advocate, find 

 at stated intervals new species, found- 

 ed upon little else than finer or 

 coarser striation, or perhaps somewhat 

 different outline. It is, no doubt, 

 quite a comfortable way of working, 

 and of keeping one's name before 

 the public when one finds what is 

 supposed to be a new diatom, if, only 

 knowing enough to distinguish the 

 genus, one measures, more or less cor- 

 rectly, the length, breadth, or diame- 

 ter, and the number of striae in .001 

 of an inch, giving sometimes a repre- 



