48 SCIENCE PROGRESS. 



cloacal diverticula. They may be regarded, at any rate 

 provisionally, as the forecasts of the cloacal chamber of the 

 future zooid, — a view which would justify the comparison I 

 have drawn above between them and the otherwise unique 

 peribranchial canals of the Pyrosoma stolon. 



It is also very difficult to accept Uljanin's account of the 

 fate of the pair of pharyngeal outgrowths. According to 

 this investigator each of these elements soon divides, and 

 one of the portions unites with its counterpart to form the 

 unpaired rudiment of the pharynx and alimentary canal of 

 the bud, while the remaining parts constitute the paired 

 forecasts of the genital gland. Analogy is altogether 

 against such a state of things as this. In no other Tunicate 

 do the pharyngeal outgrowths ( = epicardial tubes) give rise 

 to the generative glands. These invariably arise either 

 from the mesoderm, after its differentiation from the 

 endoderm, or directly from the generative gland of the 

 parent. Grobben's statement that the alimentary canal and 

 generative organs of the bud arise from parental elements 

 which are distinct from the first is much more likely to be 

 true, since it has the force of analogy in support of it. 



Be this as it may, it is quite clear that in the formation 

 of buds in the Tunicata all three germ-layers of the parent 

 participate. In the simplest cases the bud is composed of 

 the three undifferentiated layers alone (e.g., Clave linn, 

 Distaplia, Amarcecium, Perophora) ; but in the majority of 

 instances the bud consists in its earliest stages of parental 

 material which has already undergone differentiation to a 

 greater or less degree. In these cases the bud consists of 

 an aggregation of a number of organ-rudiments instead of 

 the three undifferentiated layers ; and in all satisfactorily 

 investigated cases the special elements of which the bud is 

 composed are direct prolongations or derivatives of the 

 corresponding organs of the parent. The nervous system 

 seems rarely to be prolonged into the bud in this manner, 

 but to be derived anew by delamination. Diplosoma, how- 

 ever, furnishes an exception to this rule, as the recent 

 researches of Salensky (27) show ; and Pizon has 

 endeavoured to make out the same phenomenon in the 



