1926 SYSTEMATIC 



and that if the author has not, by some means, made indication of 

 selection, then a subsequent author has the right to choose a geno- 

 lectotype out of those genosyntypes : to be precise, he should pin down 

 one definite specimen as the standard of reference. For the holotype 

 of Ab may never have been fixed ; it may be obvious from the author's 

 diagnosis of his genus that he has never seen the original material, 

 so that it is equally plain that the author is giving his generic name 

 to species erroneously identified with Ab. Or, again, if there be a 

 holotype of Ab, it may be so poor a specimen that no one would wish 

 to have it as a genotype standard, because it would give rise to so much 

 diversity of opinion that the' unfortunate A would be hunted from 

 pillar to post, being found one day with one series of characters, and 

 placed with one set of associate species ; another day with quite other 

 sets. 



This case is different from that considered in my first notes (p. 5 

 above), which is concerned with an author giving a generic name to 

 three or four distinct species, out of which he has not selected one species 

 to be a genotype. Here a subsequent writer is free to choose a geno- 

 lectotype species out of these genosyntypes ; but, again, in the geno- 

 lectotype species there may be several specimens, and, to avoid 

 confusion, the method of choosing one of these as the genolectotype 

 standard would be followed as noted above. Thus there would be one 

 rule only — whether there be several species Ab, Ac, Ad, or only a 

 series of specimens passing under one name Ab, there is freedom to 

 choose, first one species and then one sp)ecunen of that species, as the 

 genolectotype standard. 



Dr. Bather's dictum that an author should always consult the 

 holotype before giving a name is a counsel of perfection ; but it is 

 scarcely practicable. Types are scattered all over the globe. Lack of 

 health and lack of means, two interdependent afflictions of many 

 scientific men, would effectually bar the necessary travelling. 



For just such reason I hold that the material which an author has 

 under his hand, especially if he has used it in description or illustration, 

 even without having actually mentioned it as genotypical, should take 

 precedence. The genotype should go with what the author has seen, 

 and not with what he may not have seen. If the author's b turns out 

 to be a prior-named or a post-named c, then the genotype follows c, 

 not b ; but the author's specimen of c may not be the holotype of c : 

 it may be a plesiotype, or it may be a paratype. 



Therefore, with much regret, I differ from Dr. Bather's statement 

 that ' the holotype of the genotype is the ultimate standard of 

 reference.' In my opinion, the genotype standard of reference is by 

 no means necessarily the holotype of the genotype. 



If only students will in future be precise in stating which specimen 

 is to be taken as the genotype standard of reference, much trouble would 

 be saved. Unfortunately, such technical details are not impressed on 

 them by their teachers, because the teachers themselves lack the 

 necessary training. Editors, however, could do much by insisting that 

 such details be given. 



Systematic 

 Suture-line : For the distinction of genera the relative length and 

 development of the lobes of the suture-line is found to be of prime 



