430 
Transactions of the Society . 
multiplication of species on similar grounds occur in the groups con- 
taining Pinnularia nobilis, P. major , and P. viridis ; in § Asym- 
metrica of Gomphonema ; and in forms allied to Navicula aspera, to 
N. liber , to N. firma, to Bhaphoneis amphiceros , and to Triceratium 
Favus. 
The character upon -which Prof. Smith chiefly relied, regarding it 
as “ sufficiently constant to form a safe guide ” to the determination 
of species, was striation, and especially the relative fineness or coarse- 
ness of the striae. In this belief he maintained that “ striation is the 
best guide.” A more extended examination of forms from different 
localities has, however, shown that the supposed constancy of this 
character does not exist. The range of variation, so gradual that 
each step is almost imperceptible, is, on the contrary, very great. 
Striation is, in fact, only another term for cellulation, which in one 
form or other is observable in almost all diatom valves ; and although 
the relative fineness or coarseness of the cellulation of discoid forms is 
a character still frequently made use of for specific distinction, it is in 
reality one of extreme variability. Of this any one may he convinced 
by examining any gathering which contains in abundance such a 
form as Coscinodiscus concinnus variety Jonesianus. Yalves will be 
found with the cellules three times as fine as those on other valves, 
with a complete series connecting the two. In other species, as for 
example Coscinodiscus elegans , Cestodiscus pulchellus, and Melosira 
granulata , the two valves of the same frustule not unfrequently differ 
greatly in the size of their cellules. In Denticula the costae on the 
two valves of the same frustule, and in Pinnularia even those on the 
two sides of the same valve, occasionally differ considerably from each 
other. In Cocconeis the different striation of the upper and lower 
valve is now well known ; but before it had been fully recognized, it 
sometimes led to the two valves being placed in different genera. 
Thus Mastogloia maxima Grunow is nothing but the lower valve of 
a form of the common and extremely variable Cocconeis scutellum. 
In view of the magnitude of these differences in valves, which can be 
shown to belong to the same species, distinctive characters based upon 
comparatively very slight differences in the spacing of striation, in 
other words of the cellulation, are clearly inadmissible. 
Not only does the spacing of the striation vary, but occasionally 
its character also. When a frustule is in course of subdividing, the 
appearance of the striation on the two newly formed inner valves is 
sometimes entirely different to that on the two older outer valves. 
The difference obviously arises from the valves being in different 
stages of development; yet, if the older and younger valves were 
observed apart from each other, they would probably be regarded as 
distinct species, and might even be referred to quite different sections 
of the genus. 
A character dependent upon the so-called striation, which is 
equally subject to variation, is the position and extent of the blank 
