306 Prof. T. Thorell on Dr. Berthau's 



formed of but a small number of parts (segments, extremities, 

 &c.), wliich parts, again, show only slight variations in the 

 different spiders ; from this it follows that the entire group is, 

 upon the whole, of a highly uniform aspect, exhibiting but few 

 important structural points on which to rely for a natural 

 classification. 



Other difficulties arise from the fact that most of the cha- 

 racters generally found to be constant, and therefore of import- 

 ance in the classification of these animals, may yet vary most 

 materially in one and the same group. The tarsal claws, for 

 instance, the number of which (three or two) gives such 

 good and reliable characters for many tribus and families, 

 may, however, within the same family, be sometimes two, 

 sometimes three; in a few genera {PalpimanuSy Dasumia) 

 some of the legs have, in the same animal, three, and the other 

 legs only two tarsal claws. The distribution and the number 

 of the eyes, wdiich also often give sure characters both 

 for tribus and families, may nevertheless be very different 

 within the same family or even the same genus {Nesticus, 

 Haditcs). It might have been expected that, just as the 

 presence and peculiar structure of the spinnmg-ajyparatus is 

 perhaps the most salient and most characteristicfeature through- 

 out the whole Order of Spiders, so the number and the shape of 

 the spinners ought to offer reliable characteristics for the 

 different higher and lower groups within the Order ; but even 

 this is far from being the case, as I shall have occasion to 

 remark further on. 



Add to this that the two sexes of one and the same species 

 often differ from one another in the most important points, 

 and tliat the young specimens are often very unlike the adults, 

 and it must be admitted that it is not an easy task to draw 

 up a natural classification of this order of animals. 



If (passing by, for the moment, the more special criticisms 

 in Dr. Bertkau's works, viz. those which relate to the families 

 and genera, and which we shall take into consideration as 

 suitable opportunities occur) we fix our attention on his ob- 

 jections to dividing the Spiders into the seven tribus Orbite- 

 laria?, Retitelarite, Tubitelarice, Territelari^e, Laterigradge, 

 Citigradai, and Saltigradge, these objections may perhaps be 

 summarized as follows : — 



1. A higher group, suborder or tribus, is natural only on 

 the condition that all the families and genera included in it 

 are more closely related to each other tlian to any genus or 

 family of another suborder or tribus [B, p. 345) . But in the 

 system of classification in question there are genera which, 

 though belonging to one and the same family, differ more 



