190 P, HERBERT CARPENTER. 
cords (5, 6). Inthe centre of each joint of the ray, arm, 
or pinnule, the cord enlarges slightly and gives off four 
branches, or occasionally more (Pl. XII, figs. 11, 12, a’). One 
pair runs towards the dorsal surface of the arm and breaks 
up into successive subdivisions, which can be traced no 
farther. The other pair extends towards the ventral side, 
branching freely on its way. Some of its subdivisions spread 
out upon the ends of the muscular bundles, as described by 
Dr. Carpenter. This is shown in one of the first radials of 
the section represented in fig. 14. In the arms and pinnules 
other branches proceed further towards the ventral side, and 
may be traced into the small marginal leaflets bordering the 
ambulacral groove. In figures 11 and 12 on Pl. XII, I 
have embodied the results of the study of series of sections 
through grooved and ungrooved pinnules of Actinometra 
polymorpha. In this species the perisome contains a number 
of more or less regularly arranged spaces in the connective 
tissue, aud the branches of the axial cords run in the par- 
titions between them. They have no regular mode of sub- 
division, no two pinnules being exactly alike; while they 
are not symmetrical on the two sides of the same pinnule. 
Mindful of Hubrecht’s discoveries respecting the nervous 
system of the Nemertines, I have endeavoured to determine 
whether the smaller branches unite outside these spaces so 
as to form a continuous layer, but I have utterly failed in 
following them to any distance, and am quite unable to say 
how they end. Fresher specimens, which have not been 
eight years in spirit, might possibly yield more favorable 
results, but none of those which I have been able to examine 
are at all suited for very minute histological work. 
The branches of the axial cords mentioned above seem 
to have escaped the notice of the German observers, who 
either make no mention of them at all or intimate that 
they have failed to find them. They are, it is true, better 
developed in some of the large tropical Comatule than in 
the European species; but it is not very difficult to obtain 
a section of an arm or pinnule of Amtedon rosacea or 
A. celtica, in which one of the ventral branches may be 
followed all the way from the axial cord to the end of one 
of the marginal leaflets, where it becomes lost. The exis- 
tence of these widely-spreading branches requires a more 
satisfactory explanation of the character of the axial cords 
than that given by Ludwig (19, 20). He regards them as 
permanently uncalcified remnants of the connective-tissue 
basis of the skeleton, and supposes them to effect the trans- 
mission of nutritive fluid from the chambered organ into 
