270 PROCEEDINGS OF THE ACADEMY OP [1885. 



also by P. II. Carpenter, to represent a bisected proximal, and 

 the two or three plates which the}' enclose were supposed to be 

 anals or plates of the anal tube — a mistake easily explained by 

 the fad that the plates stand in line with, and join the four large 

 proximals, and have very near their size. We discovered our 

 mistake when we found that in all internal casts the radiation 

 follows I he median line of the plates, and not the suture, as in the 

 case of the proximals. The disturbance in the arrangement of 

 the two posterior radials is evidently due to the anal structures, 

 which pushed these plates out of their regular position. In 

 species with a large subcentral anal tube, the position of these 

 radials is so completely altered that they are sometimes actually 

 placed within the semicircle of the four large proximals. Such 

 is the case in the specimen of Te.leiocrinux (PI. 4, fig. 4), in which 

 the anal appendage is almost central. In this specimen, all three 

 anterior rays have three primary radials, while the two posterior 

 ones have four. The inner plates serve as a kind of axillary for 

 the ambulacra of the postero- and antero-lateral rays, which are 

 undivided for some distance, giving off" underneath a branch to 

 the outer radials. The presence of a fourth radial is rather an 

 exception, and, indeed, three radials are found, as far as we know, 

 only in the larger species of the Actinocrinidae. In species in 

 which the covering plates pass out to the surface of the vault, 

 the radial dome plates are frequently either wanting in the three 

 anterior rays, or the}' are exceedingly rudimentary and very 

 irregular in form, while those of the posterior rays are generally 

 intact (PL 7, figs. 3, 9, 10). But in some species the posterior 

 radials are partially or totally resorbed (PL 7, fig. 8), and the 

 covering plates pass out directly from beneath the central piece. 

 In Melocrinus and Gyathocrinus aluto.ceus, in which the anal 

 structures are comparatively narrow, the central piece being 

 generally surrounded by only six plates, of which two face the 

 posterior side — all five radials are placed outside the ring of 

 proximals; but we have a specimen of Melocrinus Konincki in 

 which, exceptionally, the plate of the right postero-lateral ray is 

 placed in line with the proximals. Another interesting departure 

 from the general rule is found among the larger species of Dory- 

 crinus, Megistocrinus and Agaricocrinus, in which the central 

 plate is isolated from the proximals by a belt of small pieces. 

 \<>i even the proximals are connected with the radial dome 



