of the Genera of the Aranea?. 7 



tlie ninepins so well known to and so justly feared by every- 

 one who lias endeavoured to fix genera upon solid ground, 

 because under any process where an assumption is made that 

 such a species was witlulrawn under a«oMer name and placed 

 in a new genus, and further steps based upon that action, 

 there is always the possibility that it may turn out that the two 

 s|>ecies were after all not identical, and down come several 

 ninepins, and tiie whole position has to be reconsidered. 



We have to recognize and face this possibility. What we 

 want to do, however, is to avoid as much as possible any steps 

 of eliuiination which might court such a catastroj^lie. 



'J'he case of the genus Neriene (see page 9 of the present 

 paper) will furnish a good illustration. Blackwall had a 

 perfect right to include N. marftinata if he pleased, and it 

 cannot be helped if, being the last species lelt in, owing to 

 the withdrawal of the other two, it becomes the type of 

 JS'eriene, and being congeneric (sec. autliors) with Linyphia, 

 he loses his genus as a synonym of the latter. 



Jf, however, this species be regarded as removed by impli- 

 cation under Linyphia, Latreille, then y. corniita will be left 

 in as the tj-pe of the genus. In this case cornutn would not 

 be available as the type of Dlcyphvs, Menge, as it would 

 be if " implication " weie not recognized. 



But it may afterwards be discovered that A'^. marginata is 

 not really congeneric with Linyphia, and it renews its right 

 to serve as the type of Neriene, and being the last left in 

 becomes the type : cornula forthwith loses its position as 

 the type and claims again a right to serve as the type of 

 Dicyphiis, other subsequent genera will be involved, and so 

 on to distraction. 



Apart from these lamentable consequences, retrospective 

 implication cannot be upheld, because it involves the denial of 

 the right of an author to include any species he likes in any 

 genus he likes to make, and claim that species as the type, 

 even to his own undoing. 



Subsequent withdrawals of species, moreover, can obviously 

 only take place directly under the same name or under a 

 recognized synonym. 



It is unlortunate that there appears to be no way of 

 avoiding the catastrophe I have referred to above in connexion 

 w ith steps based upon an assumption of the identity of species. 

 It might be urged, W^hy not wait until the synonymy of all 

 the species involved has been finally settled? Theoretically, 

 no doubt, this should be the first step to take; practically it 

 cannot be done without years of labour, and even then with 

 very doubtful results. What liopes can one entertain when 



