18 W. M. BALE ON SOME OF THE DISCOID DIATOMS. 



distinct differ only in having the markings a little smaller or 

 larger, while others are characterised by trifling distinctions of 

 detail which, on examination of an extended series of specimens, 

 are found to break down utterly. On the other hand it will be 

 seen that, in many instances, details which might be helpful in 

 the discrimination of species have been generally overlooked. 



The first serious attempt to grapple with the difficulties 

 involved in the classification of the genus was that of Grunow,. 

 in his work on the Diatoms of Franz-Josef Land, a perusal of 

 which leads one to regret that this acute observer did not carry 

 out a more comprehensive survey of the whole genus. Rattray's 

 Revision, though giving evidence of a vast amount of painstaking 

 research, is far from final in regard to the species admitted, 

 many of which are characterised by features obviously not of 

 specific sometimes not even of varietal value. Moreover, in 

 working over slides from well-known deposits, one finds many 

 forms which it is impossible to place under any of the species 

 described, though it is most unlikely that Rattray could have 

 failed to observe them. The impression is produced that many 

 of the descriptions have been framed on particular specimens,, 

 without any allowance for the range of variation usually present. 

 The "key" is minimised in value owing to the use in many of 

 the sections of characters which are quite inconstant, or which 

 may characterise the type but not the varieties, while the attempt 

 to include all the sections in one key has added much to the 

 difficulty of the undertaking, and has involved mistakes which 

 render it in some cases quite unreliable. (As an example, let 

 the observer take a typical valve of C. asteromphalus and attempt- 

 to trace it through the key, and he will fail to find it. But it 

 appears under Section 116, and, if followed backwards, it will 

 be referred to Section 111, where the description is, "Markings 

 rounded, granular ; interspaces hyaline, unequal, rows radial," 

 which obviously cannot apply to the species at all.) 



Nevertheless Rattray's work undoubtedly represents a great 

 advance in its suppression of a large number of pseudo-species, 

 though one cannot but regret that the process has not been 

 carried further. 



Mr. Cox, going to the opposite extreme, would reduce all the 

 multitudinous forms of Coscinodiscus to seven species, Actinocyclus 

 Ehrenbergii being included as one of them. Some diatomists 



