BRUES: NEW SPECIES OF PERIPATUS. 313 



very few in number, and the bases of the primary ones are rectangular 

 — not irregularly polygonal as in P. barbouri. A remarkable simi- 

 larity is seen, however, in the bifurcation of two transverse folds on 

 each body segment toward the middle of the dorsum. This occurs 

 in both species in exactly similar fashion. 



With P. trinidadcnsis the affinities of the Grenada species are very 

 great, as well as with P. imthurmi, P. sedgwicki, and certain forms of 

 P. dominicac. 



From P. trinidadcnsis, it may be distinguished by the entire absence 

 of dorsal lozenge-shaped color markings, and by the not very promi- 

 nent transverse tegumentary folds. There appear also to be more 

 very small accessory papillae, although middle sized ones showing 

 all gradations to the primary papillae are always present on the 

 folds, but otherwise the resemblance between the two forms is ex- 

 tremely close. Just as P. trinidadcnsis is more or less intermediate 

 between P. imthurmi from Guiana and the West Indian P. dominicae 

 with its varieties, P. barbouri must be interpolated between trinida- 

 dcnsis and dominicac. There can be no doubt, however, that the 

 similarity to the former is much greater than to the latter since the 

 gradation between the sizes of tegumentary papillae at once dis- 

 tinguishes it from P. dominicae, which falls on this account into a 

 different group in Bouvier's classification. The latter species has 

 fewer legs (29 pairs) on the average also, at least in the typical form, 

 although the variety antiguensis may possibly average 30 or 31 pairs. 



One is led to believe, therefore, so far as the evidence furnished by 

 the Grenada Peripatus is concerned, that the faunal relationship of 

 Grenada with Trinidad is much closer than with St. Vincent to the 

 north as the latter island supports Peripatus juliformis, a form that 

 extends northward with varietal variations as far as St. Thomas and 

 Jamaica, and one which is widely different from P. barbouri. 



Unfortunately no Peripatus has been, so far, reported from Tobago, 

 which lies to northeast of Trinidad and southeast of Grenada. The 

 genus undoubtedly must be represented there, and one may surmise 

 that forms identical, or at least very close to those on Trinidad will 

 be found as the fauna of Tobago is but little different from that of 

 Trinidad. 



From the foregoing it will appear that Grenada probably represents 

 the northern limit of extension of the group so well-developed in the 

 northern coastal region of South America. That these forms may 

 extend still further to one of the Leeward Islands, is possible since 

 Bouvier has described from Guadeloupe a variety bavayi which he 



