Bibliographical Notieet. 87 



innoTation in uniting P. niveut with P. tarixu ; but I feel constrained 

 to take this bold step, even at the risk of not being soon forgiven. 

 I had for a long time great misgivings on the subject." We greatlj 

 regret that his misgivings did not last longer, and can assure him 

 that he was quite justified in his fears, and that his " conchological 

 friends" are "terribly shocked" at his merging the four shellf 

 named in their allied species. It is no mere individual opinion 

 we express, but we believe it to be the opinion of all our leading 

 British conchologists, that Mr. JeiTreys has been guilty of a most 

 barbarous murder in the slaughter of these little iimocents. Are 

 not the grounds on which these species are reduced to the rank of 

 varieties untenable? It is to this general question that we shall 

 address ourselves, because the limits of a brief review do not per* 

 mit of our extending our observations to the discussion of the daimi 

 of the individual forms to specific rank. There has been an 

 axiom put forward, originating, if we mistake not, from Mr. Alder» 

 that, " if two nearly allied forms live together under the same cir« 

 cumstances, without showing any intermediate forms, the presump- 

 tion is that they arc specifically distinct." This is a sound argument. 

 But the converse of this by no means holds good. It is a most false 

 argument, that, if two ncarly|allied forms do not occur together, this 

 is to be received as proof of their specific identity. Yet this is the 

 chief ground on which Mr. Jeffreys relies in his amalgamation of the 

 above-named species. Peetem niceu* occurs throughout the Hebrides, 

 but P. variua is wholly absent from the district ; Nuatla radiata is 

 found in Milford Haven, but " always in separate parts of the baj 

 from P. nucleus." Attarte elliptiea has never been met with by the 

 author " on the same ground " with A, nUeala ; and Moetra elliptic^ 

 is regarded as a deep-water form of M. toiida. Now we are not 

 prepared to deny — very far from it — the ezistcnoe of races; but 

 mokt certainly these cannot be cited as instaaoes ci this kind of 

 variation. The allied species in question, if we except P. nitnu, are 

 found constantly in the same locality, if not actually on the same 

 ground, with the species with which Mr. Jeffreys would unite them ; 

 and they are thos associated over a considerable portion of our own 

 seas, as well as northward or southward of them. These, we repeat, 

 are no instances of races, wijich are synonymous with local varieties. 

 Had it been true that A. elliptiea, N. radiata, and M. elliptiea 

 occupied a totally different area of distribution from their allies 

 A. sulcata, N. nucleus, and AI. soltda, then such a fact might ba 

 received as an argument that the allied forms were two races of one 

 species. But this is not the fact. The cases before us are examples 

 of nearly allied species which constantly coexist in the same limited 

 area. The fact that they do not live together upon the same ground 

 and have not the same habits must surely be r^arded as an evidence 

 in favour of, rather than against, their specific rank. Another point, 

 which it appears to us that Mr. Jeffreys has lost sight of in dealing 

 with these species, is that a number of minor differences become in 

 the aggregate equal to a single more marked character. 



In the formation of genera, we find that CreneUa has been limited, 



5* 



