110 PROCEEDINGS OF THE NATIONAL JklUSEUM vol. iii 



1, any variations almost always affecting the sequence of the last two or 

 three. Modifications of taxonomic value involving leg characters are 

 those of the ventral macrosetae or spurs, and the shape of the anterior 

 male legs and their tarsal claws. 



The occurrence of large regular spurs on the ventral apices of the 

 leg joints is to be noted in most spiroboloids, although in the majority 

 of groups the number and stability is greatly reduced. Atopetholids 

 show perhaps the primitive condition in that a definite spur formula 

 holds for each species, with only occasional variations. The coxal 

 joint always bears a single seta, although in old specimens it may be 

 broken off. The number increases in going distad on the leg, and the 

 maximum number is found on the tarsal joint. Although the formula 

 is constant for a species, the formula does not follow a phylogenetic 

 arrangement, and one set of values may recur in several different 

 genera. Three different formulae are given here, to indicate their 

 occurrence: 1-1-2-2-2-6: Centrelus kerrensis (Chamberhn & Mulaik), 

 Comanchelus huhrichti, new species, and Arinolus torynophor Chamber- 

 lin; 1-1-3-3-3-6: Atopetholus angelus Chamberlain ; and 1-1-2-2-2-8: 

 Eurelus soleatus Cook. 



In nearly all the species examined for the character, the tarsal 

 joint bears a few short setae on the caudal side and usually one on 

 the cephalic face. In Onychelus obustus, however, these lateral setae 

 are much more numerous and very elongated, and form a sort of 

 fringe on each side of the leg. Conceivably this may be an aid to 

 locomotion in dry, loose sand, Onychelus being known to occur in 

 desert regions. 



In many species, the coxae of the anterior male legs are produced 

 into variously shaped lobes, the lobes of the third leg pair usually being 

 most modified. These elaborated coxal lobes seem to characterize 

 species rather than genera for the most part, and although probably 

 a specialization of a sort, do not occur in obviously specialized genera 

 such as Arinolus. The pinnacle of coxal development is reached in 

 Eurelus soleatus, where the lobes of the third legs are elongated and 

 deeply notched just before the end, the notches being engaged by the 

 coxal lobes of the fourth pair with a nice interlocking mechanism, the 

 advantage of which is difficult to im.agine. In the related species 

 Comanchelus huhrichti, none of the anterior male legs are lobed. In 

 Atopetholus angelus the coxae are produced into small blunt cones, 

 which reverse the normal sequence by increasing slightly from the 

 third to the seventh leg pairs. 



The large coxal lobes of Eurelus are accompanied by a marked re- 

 duction in the size of the tarsal claws to mere uncate remnants (figs. 

 4L,d-f) . These claws are likewise reduced in Centrelus, but are normal 

 in size in Comanchelus of the same subfamily, and in most other 



