410 Mr. F. O. P. Cambridge—A Revision 
(4) The custom of regarding as the type of a genus that 
species whose specific name has been adopted as the name 
of the genus including it has long been followed by zoologists. 
Examples of such identity of generic and specific names will 
readily occur—e. g. Lutra lutra, Scomber scomber, Astacus 
astacus, Avicularia avicularia, &e. 
The reason for such custom is clear. If a species be defi- 
nitely referred to by the author in connexion with the formation 
of a genus, it is assumed that the author himself regarded 
this as the type. How then, it is argued, can an author 
refer more pointedly to a species than by adopting its specific 
name to denote his new genus? This species, then, is regarded 
as the type and has prior claim as such over any other left by 
subsequent process of elimination or citation, just as would be 
the case if the author himself had cited the type at the time 
of founding the genus. 
But in any case the species must have been originally 
included in the genus, e. g. Avicularia avicularia (Linn.). 
The type, however, of Yarentula, Sundevall, cannot be 
T. tarentula (Fabr.), because this species was not originally 
included amongst those referred to the genus. Fortunately 
very few cases of such identity of name occur, so that compli- 
cations are not likely to arise. 
The oldest Species as the Type. 
There appear to me to be no reasonable grounds for sup- 
porting the contention that the oldest species ought to be 
regarded as the type in heterotypical genera where no type 
has been previously cited, definitely indicated, or left in by 
exhaustion. 
No one can confidently assert that authors always had the 
oldest species in their minds when they diagnosed their new 
genera, any more than that they had the jirst in order of 
publication in their minds. One might equally well assert 
that the middle species of the series would be more especially 
in the author’s mind, and that he was grouping the others 
round this central typical form. There can be no grounds 
for either of those assumptions, and no two authors would 
nowadays agree on any one of them. 
It seems to me better to leave the selection in such cases 
entirely to the option of the selector. 
There are several other points also about which we must 
make up our minds, to which I now briefly refer. 
