360 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



occurs); but no matter what the relation of the axillary is to the preceding 

 ossicles, the relation to the succeeding ossicles is always the same, and, furthermore, 

 it is always the same as the relation of the radial to the next succeeding ossicle. 



Smco axillarics are clearly most intimately related to the succeeding ossicles 

 and show no relationship whatever with those immediately preceding, it is natural 

 to infer that the same interdependence holds in the case of the morphologically 

 strictly comparable radials; that is, that the radials are in reality arm plates, and 

 are not in any way to be regarded as calyx plates, in spite of their position as an 

 integral part of the covering of the body wall. 



In certain crinoids, which have relatively enormous bodies and short arms, the 

 radial may be separated from the infrabasal beneath it by an extra plate, which 

 disappears in the later types, persistmg in many beneath the right posterior radial 

 only. The so-called "anal" of the young Antedon is the last remnant of this ])late, 

 shoved far out of its normal position. 



The radial is the equivalent of the asteroid terminal; therefore these subradial 

 plates occupy precisely the same situation as the asteroid brachials, of which they 

 appear to be the direct representatives; but they are dropped in all of the more 

 specialized crinoids, mcluding all of the recent forms, which thus show a reversion 

 to the more compact echinoid type of test, profoundly modified by the inclusion 

 in it, as a fundamental feature, of the radial, corresponding to the asteroid termmal, 

 but not corresponding as an entity to any echinoid plate. 



The occurrence of subratlials in the crinoids with large calices indicates the 

 very close connection between the radials and the brachials succeeding, strongly 

 suggesting that the radial is in reality an arm and not a calyx plate. Moreover, 

 were the ra<lial a calyx or coronal plate homologous with the ocular of the urchin 

 (a view very commoidy held), we certainly should not expect it ever separated 

 from the apical portion of the animal by subradials. 



There are only two series of true calyx plates in the crinoids — the mfrabasals 

 and the basals — corresponding to the oculars and to the genitals of the urchins. 

 The radials and all subsequent plates belong to the appendicular series and not to 

 the calyx scries at all. 



An appreciation of this fact, taken in connection with an appreciation of the 

 true interrelationships between the crinoids and the urchins, gives us a suggestion 

 as to the true phylogenetical significance of the radianal, anal x, and the interradials. 



Anal X and the uiterradials rest directly upon the basals, and thus corrospontl 

 exactly to the interambulacrals in the urchins, which follow the genitals in the 

 same way. 



Now the radials are <louble plates, the equivalent of two (or more) of the 

 ambulacrals of the urchins, and are separated from the infrabasals, the equivalent 

 of the oculars of the urchins, by the closed circlet formed by the basals. 



The radianal is occasionally (though only very rarely) interpolated in the circlet 

 of basals, so that it forms a single plate separating two adjacent basals, and connect- 

 ing the rathal with the infrabasal beneath it. 



It is thus possible to regard the interradials and anal .r as the basal ossicles 

 of the interambulacrals of the urcliins, and the radianal (including the other sub- 



