360 



The Development of Peripatus Capensis. [May 21. 



indicated below, with regard to these changes which, are still some- 

 what obscure. 



When the appendages appear, the somites become prolonged into 

 them. The portion of the somite in the leg soon (at about the end of 

 the spiral stage) becomes separated from the remainder, which shifts 

 dorsalwards, and comes into contact with its fellow along the dorsal 

 middle line above the gut. I will in future refer to the former as the 

 leg division of the somites, and to the latter as the dorsal division. 



The leg division becomes inconspicuous in the antennas and jaws. 

 In the remaining appendages (oral papillas and seventeen pairs of 

 legs) it acquires an opening to the exterior at the base of each 

 appendage and immediately outside the nerve cords. 



The portion of the leg body cavity, which thus communicates with 

 the exterior, gives rise to the segmental organs. The segmental 

 organs are never completely separated from the rest of the leg body 

 cavity, except in the case of the first, the aperture of communication 

 persisting as the internal opening of the nephridium. 



The anterior segmental organ, viz., that of the oral papilla, gives 

 rise to the salivary glands of the adult.* 



The segmental organs then develop in Peripatus, as they do in 

 Elasmobranchs, by the direct modification of parts of the 'primitive body 

 cavity. These constitute, I believe, the only two cases in which such 

 a method of development of these organs has been observed. 



The cells of the splanchnic lining of these dorsal divisions of the 

 body cavity contain during the spiral stage the primitive generative 

 cells. But it is difficult to say where the generative cells originate, 

 because during the spiral and later stages nuclei exactly similar to 

 those of these cells are placed in groups in the portion of the endo- 

 derm which is in contact with that part of the splanchnic mesoderm 

 which contains the generative cells. I cannot therefore be certain 

 that the generative cells do not arise in the endoderm and subse- 

 quently migrate into the splanchnic mesoderm. However that may 

 be, they soon become surrounded by some of the smaller flat cells of 

 the mesoderm which constitute the egg follicle. So far the develop- 

 ment of the somites is quite clear and easy to follow. But the 

 changes by which the dorsal divisions of the somites are converted 

 into their permanent form take place at a late period of development 

 — during November — and are, in consequence of the thinness of the 

 walls, extremely difficult to follow. I have not succeeded entirely in 

 following them. 



I will content myself, therefore, with making the following state- 

 ment, of the truth of which I am by no means confident. 



The dorsal divisions unite with each other transversely and longi- 



* The opening marked d in fig. 8, Plate 75, of Moseley's paper {op. cit.) is the 

 opening of the salivary gland, not of the slime gland as he states (see below). 



