THE MICROSCOPE. 135 
cases the spines and similar processes, where well developed, are 
important factors in the growth of the diatom, and, where obseure 
or minute, are lingering traces of a feature being eliminated by the 
process of evolution. 
The spinous discoid forms are mostly found in the third tribe, 
Crypto-raphidiee, of Prof. H. L. Smith’s classification, followed by 
Van Huerck, in which tribe, however, the absence of a true raphe is 
an essential characteristic. In the first tribe, the Raphidiez, the 
absence of spines is equally an essential character, so that either an 
entire new tribe and family, as well as genus, must be erected to 
receive Raphidodiscus, or else the character of Tribe I or Tribe IIL 
must be slightly modified to admit this form. So far as I know, this 
diatom is the only one which does not fall naturally and easily into 
one or the other of the three tribes of Prof. Smith’s admirable 
classification, and it would seem far preferable to modify the 
characters of one of the existing tribes enough to receive this form 
than to introduce a new tribe of so limited scope as to include but a 
single family, genus and species; for such, I believe, would be the 
result of that course. If one of the existing tribes is to be modi- 
fied, it should be the one containing most of the features of the new 
form and its nearest allies, and this is obviously Tribe III, which 
already contains all the features of the new valve, except the raphe. 
I suggest, therefore, that the negative characteristics of the 
Crypto-Raphidiex, which now stands as “never with a central 
linear blank space or true raphe on the valves,”’ be so modified as to 
read “without a central linear blank space or true raphe on the 
valve, except a raphe or pseudo-raphe in Raphidodiscus.”’ 
From examination of a number of specimens, I feel convinced 
that the three described species of Raphidodiscus must be reduced 
to a single one. The impression conveyed to everyone by the 
photograph of R. Marylandica, that it was a case of two different 
diatoms accidentally wedged together, is confirmed by the examina- 
tion of specimens, as the markings of Melosira can be plainly made 
out when those of Raphidodiscus are out of focus, and vice versa. 
R. Christianii is simply R. Marylandica without the enveloping 
Melosira; or, more correctly, Marylandica, is Christianii lodged in 
a Melosira. R. Febigeriiis simply Christianii with its marginal rim 
broken away. In a slide belonging to Dr. D. B. Ward, of 
Poughkeepsie, N. Y., containing six valves of Raphidodiscus, one is 
the so-called “‘ Marylandica,”’ pronounced genuine by Mr. Christian ; 
four are the typical “‘Christianii,” and one is the typical “ Febi- 
gerii.”” The valve of ‘ Marylandica” can be clearly seen to bea 
