66 ON A TORNARIA FODND IN BRITISH SEAS. 



througli tlie region of the proboscis pore in a perfect Tornaria. In 

 fig. 6 two canals may be seen lying side by side. Of these, one 

 marked ant. h. c. is the canal of the proboscis cavity leading to the 

 proboscis pore. The other, marked v., is the heart in question. An 

 examination of figs. 7 and 8 shows that it is formed as an invagina- 

 tion of the ectoderm just above and to one side of the proboscis 

 pore {d. p.). Prof. Spengel, who has investig*ated the whole 

 development of many species of Tornai'ia, has been good enough to 

 send me proofs of the illustrations of his forthcoming monograph 

 on Balanoglossus, and from his figures and from the account given 

 in his preliminary paper {d) it appears that this vesicle is destined 

 to form the " heart " in the adult animal, the proboscis sac of 

 Bateson. 



By permission of Mr. Weldon I reproduce a drawing (fig. 19) of 

 the proboscis gland in a later stage of the Tornai-ia found by him 

 in the Bahamas. In this the heart or proboscis sac {v.) is seen as a 

 sac lying in the proboscis gland. It is completely closed, and does 

 not communicate with the blood system nor with the proboscis 

 cavity. The most plausible explanation of this structure is that the 

 anterior body-cavity of Balanoglossus may primitively have been 

 a paired structure, and that this sac may be a member of the pair 

 and the degenerated fellow of the proboscis cavity. But the 

 development of a mesoblastic pouch as an invagination of the ecto- 

 derm lands us in a great morphological difficulty. 



The development of the posterior mesoblastic pouches can be 

 followed up to a certain point in this stage. It has already been 

 shown by Metschnikoff and Agassiz that a plate is budded off from 

 the posterior region of the gut on each side, which subsequently 

 undergoes division, and forms, as shown by Spengel, the collar- 

 cavities and body-cavities of the adult. My sections show that 

 right and left of the body a plate of cells is budded off from the 

 upper edge of hind-gut on either side. At first continuous with the 

 hind-gut (fig. 14), each plate subsequently separates from it, and 

 becomes so closely applied to the mid-gut as to look as if it had 

 originated from it (fig. 15). The cells composing these plates 

 multiply, and a cavity is formed in them, as shown in fig. 16. No 

 later stages were observed in our Tornaria, but in other forms these 

 pouches become each divided into an anterior and a posterior 

 portion, the anterior moiety on each side giving rise to the collar- 

 cavity of the adult, the posterior moiety to the general body-cavity. 

 This further development is figured in the drawings which Prof. 

 Spengel has sent me, and I was able to follow it in the series of 

 preparations of the Bahamas Tornaria lent me by Mr. Weldon. 



Bateson, in his account of the development of Balanoglossus 



