Classification of the Spiders. 323 



more increasing splitting up of the older families into numerous 

 new groups of tlie same denomination, it may perhaps not be 

 inappropriate to regard the presence or absence of the cribel- 

 lum and calamistrum as a character sufficient to distinguish 

 families j and it must then be admitted that all the cribellate 

 families adopted bjBertkau are good systematic units, though 

 I, for mj part, should prefer to unite his Dictjnida3 and Amau- 

 robiadffi in one family, Diet jnoidse, these two groups being only 

 distinguished by the difterent development of their tracheaj. 

 New as to the distribution, among the generally received tribus, 

 of Bertkau's nine cribellate families, they must be referred 

 partly to the Orbitelarise, partly to the Tubitelaria3 ; as yet there 

 is no example of a cribellate spider belonging to any of the 

 other tribus. To the Tubitelariee belong the Zoropseoidas, which 

 are closely allied to the Drassoidaj ; the Dictynoidai (inclusive 

 of the Amaurobiadaj, Bertk.), which are nearly related to 

 the Agalenoidai ; the Eresoidffi, which, though very peculiar, 

 may, as Bertkau thinks, be placed in the neighbourhood of 

 his Amaurobiadse ; the CEcobioidai, which appear to have 

 their nearest allies in the Urocteoid^ ; and probably also the 

 Eilistatoidee, which among the Cribellatee are completely 

 isolated, and have their allies among the Ecribellatte, approxi- 

 mating in some respects to the Drassoida and the Scytodoida?, 

 and even to the Territelarias. — There remain to be taken into 

 consideration the Dinopoidaj, Miagrammopoidaj, and Ulobo- 

 roidae. The Dinopoidaj, whose systematic position has been 

 so contested, and which 1 had formerly placed in the neigh- 

 bourhood of the Agalenoida3, would seem, on the strength of 

 the important reasons alleged by Bertkau {B, p. 353 et seq.), to 

 have their nearest allies in the Miagrammopoidaj and Ulobo- 

 roida3 ; as an additional reason for assigning this place to this 

 family may be adduced the presence (at least in Dinopis 

 camelus, Thor.) of so-called accessory or auxiliary tarsal 

 claws, which, so far as I know, have only been observed in 

 the Orbitelariaj and in part of the E,etitelaria3. That the 

 Miagrammopoidffi are allied to the Uloboroidffi is generally ad- 

 mitted. It therefore only remains to show that the Uloboroidaj 

 should be placed in the tribus Orbitelariie ; for if this is settled, 

 the two last-named families will, of course, follow along with 

 them. Now it is in the first place a fact (which Bertkau, 

 however, appears to doubt) that Ulohorus is a true round-web 

 spider*; I have myself captured both U. Walckenaerii and 

 tl.plumipes in their circular, perfectly closed webs; and this 

 fact is, 1 believe, one of the strongest proofs of the artificial 



* See for instance Thorell, " Till kaunedomen om slagteoa Mithras och 

 Uloboriis," in ffifversigt af K. Veteuskaps-Akadeniieus Forhandliugar, xs, 

 (1858), p. 194. 



