Classification of the Spiders. 309 



are good systematic units, even Dr. Bertkau himself would 

 seem to admit. There remain then to be discussed the Tubi- 

 telarife and the Orbitelariaj. As to the former of these tribus, 

 Dr. Bertkau enunciates nearly the same opinion about its 

 nucleus, the family Drassoidee, as I had expressed about the 

 tribus Tubitelariaj in general, viz. that in their habitus and 

 in their way of life the members of this family show a certain 

 polymorphism and manifold points of contact (" Ankliinge ") 

 with other families [A^ p. 375 ; conf. Thor., On Europ, Spid. 

 pp. 41 and 109). Just as the family Drassoidas is a natural 

 group notwithstanding its being looser and more poly- 

 morphous than most, if not all, other spider families, so the 

 tribus to which the Drassoida? belong, and which is, as it 

 were, an enlargement or amplification of that family, is, I 

 think, a natural grouj), although it be less compact and more 

 polymorphous than the other tribus. As to the Agalenoidee, 

 they are so nearly related with the Drassoid^e, and show such 

 gradual transitions to this latter family, that arachnologists 

 have, in general, had recourse to the character (in this case 

 quite artificial) afforded by the different number of the tarsal 

 claws, in order to be able to distinguish these two families ; 

 so that genera {Agroeca, for instance) which in all other re- 

 spects closely agree with the AgalenoidiB have, on the sti-ength 

 of that character, been removed from this family and placed 

 among the Drassoida3. That the Dysderoida3 (of which we 

 shall speak more in detail further on) differ from the other 

 Tubitelarite in a few important points and show some affinity 

 with tlie Territelariffi is true ; but they are, at all events, 

 much more closely related to the typical Tubitelarise than to 

 any other spiders. If the Tubitelari^e should be resolved into 

 two or more tribus, then the Dysderoidte might, of course, be 

 made to form a particular tribus^ as might perhaps also be 

 the case with the Filistatoidaa ; I for my part prefer, however, 

 for the present not to increase the number of the tribus 

 generally admitted, and think it is better to add the two 

 above-named (and other) more or less aberrant families to 

 those tribus within which they have their nearest allies. It 

 is indeed quite easy to dismember and divide the different 

 groups, tribus, families, and genera almost ad infinitum • 

 but it is more difficult and, I think, more meritorious to try to 

 unite them into higher units, and thus to form of all these 

 apparently " disjecta membra " an organic whole — a si/stem. 



2. With regard to the objection made by Dr. Bertkau under 

 this head (2), it would indeed appear as if the claim to a 

 natural classification, which he sets forth, were quite reason- 

 able J but in reality it is not so. It is not the zoologist or 



