54 Mr. F. 0. P. Cambridge — A Revision 



the original generic group to that to which it belongs by 

 implication. 



On no account must all the species be removed from the 

 title originally given to them ; one at least must be left in, 

 which in that case becomes the type. 



Where a generic name has been preoccupied, the loss of 

 the name does not lessen the value of the group selected, so 

 that a type may be selected for that group and another 

 name given to it. 



These are the main features of the process, and I here 

 give an instance to show more clearly how it works out in 

 practice. 



For instance, the name Bomhastes is given to a group of 

 three species A, B, and C, which are the only ones originally 

 included under that name by the author of it. The question 

 is, which species must we regard as the type ? 



There are two processes by which the type can be deter- 

 mined, either (i) by definite selection or (ii) by elimination. 



And both processes may be utilized in a selection. Under 

 the first process (a) the author himself may select A, B, or 

 C as the type ; or {h) another author may select A, B, or C 

 as th.e ty^e; and the species so selected must be regarded as 

 the type and no other. No author, of course, not even the 

 originator of the genus himself, can definitely select as the 

 type a species already removed either definitely or by im- 

 plication under another generic name. If he has done so, 

 his selection becomes null and void, because he had no power 

 or right to make such a selection. 



Under the second process, where no type has been defi- 

 nitely selected, one or two, but not all, of these species may 

 be removed and placed under another generic name by any 

 other author, thus " breaking up " the original genus and 

 " limiting " the generic name to one, or two, species ; the 

 last species left in being the type. 



If 13 andC are removed, A is left in and must be regarded 

 as the type ; if A and C are removed, B remains as the type; 

 if A and B are removed, C is left as the type. If A be 

 removed alone, then B or C can become the type either by 

 definite selection or by a further removal of one of them. If 

 B or C be removed, then the same remark applies to A, C 

 or A, B respectively. Ihe species left in are sometimes 

 termed the " residual sjyeciesy 



It will be evident that the settlement is comparatively 

 easy when any definite selection of the type has been made 

 soon after the founding of the genus ; but the matter becomes 

 much more complicated when the genus, perhaps involving 



