96 



The main-rays are generally short and bear 4 or 5, considerably divergent 

 branch-rays. Such spicules usually have a diameter of not more than 50 ^. 

 The terminations of the branch-rays of all these hexasters, the stoutest as well 

 as the most slender, are either all simply pointed or all bear a verticil of fine 

 claws usually four in number. These are vertical to the branch-ray or extend 

 obliquely upwards and are either straight or slightly recurved. The size of 

 these claws is subject to great variation, some are so short as to be barely 

 visible under the highest powers of the microscope, whilst the largest attain a 

 length of 3 or even 4h- . In one and the same spicule all the claws are similar 

 in shape and of nearly equal size (pi. XVI, f. 9,10). It is not at all probable 

 that these terminal claws are senile structures, produced in all hexasters when 

 they attain a certain age, because they are met with similarly developed and just 

 as frequently, if not more frequently, on the most slender and therefore probably 

 youngest, as on the most robust and therefore probably oldest, hexasters. The 

 elongated hemihexasters, in which only two opposite main-rays are differentiated 

 and terminally divided into four, five, or more, diverging branch-rays, and the 

 four other lateral main-rays I'emain undivided, are just as numerous as the " regu- 

 lar " hexasters described above. The length of their axial main-rays is exceed- 

 ingly variable and one of them is usually much shorter than the other which latter 

 is not infrequently shortened to such an extent, that its branch -rays appear nearly 

 sessile and radiate from the same centre as the other main-rays (pi. XVI, f. 8). 

 The branch-i-ays of the axial, differentiated main-rays are 20-25 h- , the simple, 

 lateral main-rays 20 f* long. Also among these spicules slender and robust, 

 oxyhexactrose and onychastrose forms are met with. More rarely spicules of 

 this kind occur, in which one or all of the lateral main-rays are terminally 

 divided into branch-rays (pi. XVI, f. 7,10, 11), I wish to add that I have 

 observed these variations of the parenchymal hexasters not only in the Indoceanic 

 "Investigator" specimens but also in the numerous other specimens examined 

 by me which were partly collected by the "Challenger" expedition and which I 

 partly obtained from other sources. 



Most of these " Investigator " specimens of Aphrocallistes bocagei Perc. 

 Wright were found in the Bay of Bengal near the Andamans, some also to the 

 south of Bombay near the Angrias bank. All were obtained in depths of 200- 

 500 m.=105-263 fths. 



As mentioned above, these si^ecimens represent two rather different forms. 

 One form is more robust and has radial diverticula of the thickness of the little 

 finger. This corresponds to the type specimen originally described by Perc. 

 Wright. The other form, represented in figs. 1 and 2 on plate XVI, is more 

 graceful and has radial diverticula only 3-5 mm. thick. I have not found that 

 the specimens belonging to these two forms are locally distinct. Both were 

 captured in most localities in nearly equal numbers, and they are, as stated above, 

 connected with each other by occasional transitions. 



