20 PROCEEDINGS OF THE NATIONAL MUSEUM vol.78 



The first two differences are difficult to evaluate. Various species 

 of Peripatopsidae approach nearer to Aysheaia in these respects 

 more than do the Peripatidae, but there is no evidence that the Peri- 

 patopsidae are primitive in this. From a general point of view it is 

 more reasonable to suppose that the small number of segments, if 

 not that of the annuli, is a specialized reduction rather than a primi- 

 tive character in Aysheaia. The greater number of claws may be 

 reasonably regarded as primitive and so may the internal appendage 

 on the legs, particularly if it is correlated with a marine habitat or 

 is homologous wdth the coxal vesicles which are found only in primi- 

 tive Onychophora and tend to undergo a progressive reduction. If 

 the interpretation of the head given below be regarded. as correct 

 Aysheaia is much more primitive in this respect than any living 

 Onychophoran. 



Head of Aysheaia. — In the recent Onychophora there are three ap- 

 pendages anterior to the legs which, being considerably modified, may 

 be considered as defining the head in the adult. These three ap- 

 pendages, the antennae, the jaws, and the slime papillae, are usually 

 regarded, following the classical work of Sedgwick (1885-8) as be- 

 longing to the first three mesoblastic somites, it being supposed that 

 no reduction in the segmentation has occurred at the anterior end of 

 the animal. The somites from which these appendages develop are 

 originally all postoral, the first later moves forward to form the 

 preoral lobes from which the antennae arise. The second and third 

 appendages, which are postoral throughout development, show cer- 

 tain characters Avhich make it reasonably probable that they are de- 

 rived from legs of the same type as the trunk legs of the adult. Thus 

 the jaws are comparable to legs in which the main axis is very reduced 

 and the claws hypertrophied, while the slime papillae may be re- 

 garded as footless legs with greatly hypertrophied crural glands. 

 The antennae differ from the succeeding appendages in exhibiting no 

 trace of an origin from legs and in originating more dorsally than 

 any of the rest of the series. Holmgren (1916) in his great work on 

 the Arthropod brain, homologises with great certainty the antennae 

 of the Onychophora with the palps of the Polychaets. Lying in 

 front of the palps in the Polychaets are prostomial " antennae " or 

 " tentacles," which are variously developed and may be median or 

 paired. These tentacles are innervated by a nerve which has the 

 same relations to the brain as the nervios tegumentarius in the Ony- 

 chophora. In the embryos of some Peripatidae, as Peripatus ed- 

 wardsi Blanchard, described by Kennel (1885) and Eoperipatus 

 weldoni Evans investigated by Evans (1901), indications are found of 

 frontal organs which presumably represent the annelid tentacles, lost 



