206 WALKER-ARNOTT, ON ARACHNOIDISCUS, ETC. 
usually called the entire frustule” consists of “two frustules 
in the act of self-division,” and consequently of four valves, 
and it results from this that what is commonly called a 
single frustule is composed of two new valves, forming or 
formed, in addition to the two primary ones ; in short, that 
the small portion seen in front alone belongs to the original 
frustule. If we now take what is usually called a double 
frustule (as in Smith’s fig. 28 d), which by Dr. Gregory’s 
theory must consist of four frustules or eight valves, three- 
fourths of the whole must be either in the transition state, or 
represent six valves, each identical with the two original ones, 
but in a different position. As I cannot suppose that Dr. 
Gregory intended to take distinctive marks from the connect- 
ing zone in its variable state, or from more valves than one, 
or from their accidental relative position, it appears to me 
that his hypothesis, if correct, would much increase the 
difficulty of finding constant characters, and cause the rejec- 
tion of several on which he depends, as well as the annihila- 
tion of many of his new forms. 
I may here mention that A. marina of Smith, alluded to 
by Dr. Gregory under his No. 76, is precisely what Dr. 
Gregory calls 4. Proteus. Smith’s figure (‘ Ann. of Nat. Hist.’ 
for 1857, vol. xix, tab. i, f. 2) is far from good, and represents 
the double frustule towards the close of the self-dividing 
process. Some years ago Smith gave it the MS. provisional 
name of A. Scotica, from its having been first detected on 
the west coast of Scotland by Mr. Hennedy of Glasgow ; he 
omitted it in the second volume of his ‘ Synopsis,’ being not 
quite satisfied with its claims to be specifically distinguished 
from A. affinis ; but these doubts were removed by afterwards 
finding it in the summer of 1856 near Havre and Biarritz, on 
the French coast, and thus having an opportunity of study- 
ing it in the living state, and drawing up a specific character. 
