MONOGEAPH OF THE EXISTING CRINOIDS. 345 



the pentacriiiitcs (fig. 566, pi. 7), or entirely losing thoii- identity and merging with 

 the topmost eolumnal, as in the comatulids and in various other forms. 



This left the basals to form the floor and the lower part of the sides of the 

 calyx, as we see in such forms as Calamocrinus or Ptilocrinus. But now the basals 

 began to undergo the same change; they became reduceil in size, and reclined to 

 a horizontal position, at the same time movmg inward over the hmer (now upper 

 or ventral) surface of the infrabasals toward the center. The basals of the penta- 

 crinites are at this stage, but those of the comatulids have gone still further, become 

 quite small and functionless, and been metamorphosed into the rosette, as already 

 explamed, excepting only in the genus Atelecrinus, where, although there are no 

 undoubted infrabasals in the adult, the basals have transformed only to the stage 

 at which we find them in the pentacrinites. 



The metamorphosis of the orals is exactly correlated with that of the liasals; 

 but it is entirely confined to resorption, so that, as the basals become reduced and 

 transformed into the rosette, the orals gradually disappear. 



In a few types, especially within the family Bourgueticrinidffi, the metamor- 

 phosis of the basals has followed somewhat different lines. Instead of gradually 

 leaning outward with the progressive development of the calyx, they have grad- 

 ually leaned inward, so that finally they have come mto a position more or less 

 parallel with the dorsoventral axis of the animal, eventually fusing and forming a 

 small and solid ring-like, cylindrical, or truncated conical calcareous element, which 

 to all intents and purposes is simply a topmost eolumnal firmly attached to the 

 radials. During this change the basals may become much reduced in size or may 

 become very greatly elongated, so that each presents a maximum surface for 

 attacliment to its fellows on either side. Various stages of this process are seen in 

 Democrinus (fig. 133, p. 203), Bythocrinus (fig. 131, p. 203), Mormchoerinus (fig. 132, 

 p. 203) and Bhizocrinus, Baihycrinus, and Rycrinus, while in Naumachocrinus (fig. 130, 

 p. 203) it is shown in its most perfected form. 



Finally the radials, both in the comatulids and in the pentacrinites, originally 

 lying in five planes each parallel to the dorsoventral axis, have gradually leaned 

 outward to a nearly or quite horizontal position, and have moved inward over the 

 inner (now upper or ventral) surface of the basals so that, properly speaking, they 

 form the floor of the calyx, and not the sides as formerly, their chief function being 

 to serve as the attachment for the arms, instead of as formerly (and at present in 

 such genera as Calamocrinus, Thalassocrinus (fig. 145, p. 209), Uyocrinus, Gephyro- 

 crinus, Ptilocrinus (fig. 144, p. 207), etc.), to protect the internal organs. 



This change in the size and in the interrelationships of the primitive calj'x 

 plates is to be accounted for solely by the gradual change in the mechanics of the 

 organisms. A globular body covered with large equal plates, just in apposition 

 at their borders and without overlap, is well suited for a pelagic existence, and we 

 see it retained only in the pelagic sj)ecie8, where it is best shown, ])robably in an 

 exaggeration of the primitive condition, in the aberrant comatulid Marsuintca 

 (fig. 565, pi. 7). 



79146°— Bull. 82—15 23 



