Notes on the morpliology of the Tuiiicata. 531 



we find among the Styelinae, Styela with a normal duct, Pohjcarpa 

 with a mere rudiment; among the CyntMinae some species of Cynthia 

 show an immense and highly glandular duct (e. g. Cynthia papillosa) 

 while in others the duct is very short (e. g. Cynthia carnca); in my 

 only example of the Bolteniinae no trace of a rapheal duct is found. 

 None of the Molgulidae show a normal rapheal duct, though in Herd- 

 mania and 3îolgula, we do find a rudiment. In Amaroecium and 

 Fragaroides alone, among the compound Ascidians, have I found a 

 trace of this organ and in these it is but a short process from the 

 gland soon losing its lumen. 



But the matter needs a little closer scrutiny. There is, I believe, 

 evidence of a somewhat closer relation between the rapheal duct and 

 the rapheal nerve than would be indicated merely by their position 

 side by side in the raphe. In many species of both simple and 

 compound Ascidians we find the ganglion and neural gland are fused 

 together. This point of fusion usually lies near the posterior ends of 

 the gland and ganglion and it is from this area of fusion, in many 

 species, that the gangliated rapheal nerve arises. Now the rapheal 

 duct is often in intimate association with this same region. Observe 

 the relations in Amaroecium (Plate 37, Fig. 47). The rapheal duct 

 runs out a short distance, loses its lumen, and becomes united to a 

 mass of cells that push out from the ganglion. This union is so in- 

 timate that one cannot say whether it is the ganglion or the rudi- 

 mentary duct which is continued into the raphe as the cord of cells 

 which in this species we find associated with the fibres of the rapheal 

 nerve. 



Again in Ascidia atra (Plate 34, Fig. 4) we find a fusion of the 

 rapheal duct with a definite cord of cells arising from the back of 

 the ganglion. The common cord thus formed soon loses its definite 

 contour and becomes transformed into a loose mass of cells, evidently 

 ganglionic, which become associated with the fibres of the rapheal 

 nerve. In Phallnsia manimillata (Plate 34, Fig. 5) we have almost 

 identical conditions, except that there is in this species no cord of 

 cells arising from the brain , or at least only a mere rudiment that 

 has no connection with the rapheal duct. The rapheal duct pushes 

 back alone, then bends downward into the raphe where it loses its 

 lumen and its cells become loosely arranged and soon unite in such a 

 way with the fibres of the rapheal nerve that we must call them 

 ganglionic in their nature. 



We can, therefore, say that in some species the ganglion cells 



