OPHIUKANS OF THE PHILIPPINE AND ADJACENT WATERS. 271 



over the rest of the disk, and their outlines are scarcely evident (pi. 

 53, fig. 5; pi. 54, fig. 3) ; sometimes the club spines are less numerous 

 (pi. 54, figs. 1, 2 ; pi. 55, figs. 3, 4) , and in this case the radial shields 

 usually become very large, as may be seen in the last example ; in this 

 also the variations may be very considerable. 



The microsopic examination of isolated club spines from the dorsal 

 surface of the disk shows that the spinules which surround them 

 have not such equal dimensions nor so regular an arrangement as 

 would be supposed from the examination of the dorsal surface of 

 the entire animal. I give (pi. 102, figs. 2a and &) photographs of a 

 rather large number of club spines from different individuals, and it 

 may be seen on some of them that the spinules of the same club spine 

 are not always of the same length and of the same diameter. These 

 differences do not appear when the dorsal surface of the disk, cov- 

 ered with club spines which are always very closely crowded and of 

 which the spinules are more or less interlocked, is examined, even 

 under the microscope; the general appearance in direct view gives 

 to the club spines a regularly stellate appearance to which the species 

 owes its name, and which is similar to that figured by Doderlein 

 {'96, pi. 14, fig. 9a). 



The study of the true spines which are sometimes mixed with the 

 club spines in 0. stelligera has brought out in a certain number of 

 them a very curious and quite unexpected character. These spines 

 are sometimes bifurcated, and I find on the same specimen spines 

 which remain simple and others which are bifurcated. These latter 

 are always less numerous than the others, and moreover they are 

 not found on all individuals. I give photographs of some of these 

 bifurcated spines (pi. 102, fig. 2/). It is seen that the bifurcation 

 occurs always very near the base, and the two branches, which are 

 parallel to each other, remain very close together. Both of them 

 reach almost the same length, and the variations which are found 

 in this respect are insignificant. The length of the bifurcated spines 

 is the same as that of the simple spines, and the structure is quite 

 identical in the two cases, the same general appearance, the same 

 denticulations, etc. The simple spines (pi. 102, fig. 2e) are ordi- 

 narily elongated in a perfectly straight line, but some of them show 

 a rather sharp curvature near their base; this curvature occurs 

 almost at the point where the bifurcation takes place in the bifur- 

 cated spines. It might be supposed that the spine belonged to a 

 group of two and that it was separated from its fellow, but that 

 certainly is not the case, for there is never seen any trace of that 

 separation and the width remains the same proximal and distal to 

 the inflection. This form is especially well seen on the second and 

 third spines to the left which I show in fig. 2e. 



