445 



It is limited dorsally only by the nerve-ganglion (ectoderm). I have 

 traced the one into the other in both the transverse sections which I 

 figured and in longitudinal sections. I should not have felt justified in 

 tinting both organs red and thus running the risk of perpetuating a 

 possible error, unless I hed convinced myself beyond a doubt of their 

 continuity as two parts of the blood-system. 



In the buds, with only one pair of tentacles, the relationship of 

 these parts is much more clearly seen. The sub-neural sinus is sphe- 

 rical and of large dimensions. A section of the region shews almost 

 diagramatically that it is formed as a haemocoele-space between the 

 ectoderm and mesoderm. At this stage there is no sign of the cavity 

 which Mr. Harmer would homologise with the 'heart' oiBalanoglossus. 

 In the adult, the communication between this space and the anterior 

 body-cavity is easy of demonstration in longitudinal sections. It gives 

 no indication of having been artificially produced. From a comparison 

 of the blood-vessels in the living Actinotroclia and their appearance in 

 sections, I am convinced that the true extent and size of the vessels of 

 Cephalocliscus are not shewn in sections, and that the sub-neural gland 

 of Cephalocliscus, like that oi Actinotroclia^ is surrounded by a blood- 

 sinus on all sides, contained between the walls of the coelomic cavities. 



Since Mr. H armor's opinions have been expressed I have natu- 

 rally been carefully over the sections of half-dozen specimens, but 

 considering that, as already mentioned, I at first held these views my- 

 self, I had already tested them and had formed them incompatible with 

 the facts of the case. 



It is with great diffidence that I thus find myself in disagreement 

 with one to whom I am indebted for my first knowledge of this spe- 

 cies, and for whose opinions I have a profound respect. 



Mr. Harmer's diagrams and my own agree in shewing that his 

 'notochord' lies essentially between the anterior body-cavity and the 

 collar-cavities and further forward than the collar-cavities (fig. 1); it 

 lies dorsal to the anterior body-cavity. 



In Balanoglossus^ on the other hand, the mutual relationships of 

 the parts in the proboscis are clearly seen in a series of transverse sec- 

 tions through a young specimen, as figured by Morgan (Journal of 

 Morphology Vol. V. Plate XXVIII, figs. 59, 60, 63, 61, 62). Here 

 it is evident (Fig. 2) that the 'Eicheldarm' of Balaiioglossus lies essen- 

 tially in a space between the anterior body-cavity and the ven- 

 tral wall of the proboscis. If, as in many species figured by Prof, 

 S pen gel, the 'Eicheldarm' moves dorsalwards in later life, its line 

 of displacement is clearly indicated by the formation of a mesentery, 

 or 'ventrales Eichelseptum' (cf. fig. 4). 



