WALKER-ARNOTT, ON ARACHNOIDISCUS, ETC, 205 
silex and imperfectly seen, then they become more and more 
conspicuous and siliceous, rapidly altermg in shape and_posi- 
tion ; the primary valves also alter in position, but not im 
form. The central nodules of the two new valves appear 
now at the back, nearly where the connecting zone formerly 
was, while the back of each half of the double frustule is re- 
moved to the side, or is placed right and left of the spectator’s 
eye, instead of directly before him; thus each half of the 
twin frustule occupies a position at right angles to what the 
parent frustule did, and which the perfect separated one 
ought to do. Finally, the two portions of the double frus- 
tule separate from each other in front like the two valves of 
a mussel-shell ; the amount of separation depending on the 
strength of the remains of the siliceous band that connects 
them, and causing the general outline to vary extremely. 
In characterising or figuring the forms of this genus, it 1s 
therefore necessary, for the sake of comparison, that each be 
viewed under precisely the same circumstances ; moreover, so 
many changes take place whilst the process of self-division 
goes on, that no descriptions or figures are of much use in 
the identification of permanent forms or species, unless when 
taken from the simple but perfect frustule before self-divi- 
sion commences, or from its valves. 
Although the structure of the genus adopted by Kutzing 
and Smith be not what I conceive the correct one, no practical 
inconvenience arises, as their specific characters have been 
selected from the same parts on which they must otherwise have 
relied. It is, however, somewhat different with that given by 
the late Dr. Gregory in his paper in vol. xxi of the “Trans. of the 
Royal Soe. of Edinb.’ p. 510; but here I experience consider- 
able difficulty from his having introduced into his descriptions 
several terms, neither employed by any other writer, nor ex- 
plained by himself, the meaning of which the reader is left 
to discover. From a careful comparison of the figures, I 
presume that by “ventral margin” he intends sometimes 
both the anterior and posterior margin of the folded valve, 
and sometimes the irregular margin of a portion of the still 
adhering connecting membrane; by “ dorsal margin,” the 
outline of the frustule; and by “inner curve line,” the 
median line of the valve ; most of which vary in appearance 
according to the position presented to the eye by the frustule 
during self-division. There are other terms, however, about 
which I am more doubtful; as “inner margin,” “outer 
margin,” and inner or outer “‘ compartments of the valves ;” 
although it is probable that all, im some way, refer to the 
connecting zone. Dr. Gregory’s theory is, that “ what is 
