344 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



tioii, binding the ossicles together far more tightly than the few large sutures of 

 Marsupites, j'et admitting of at least as much flexibility of the body wall. The 

 difference between Marsupites and Uintacrinus is found to be, when analyzed, 

 merely a difference in arm length; the structure of the arms in the two genera is 

 exactly the same; the result of the great length of the arms in Vintacrinus has 

 been to decrease the size of the calyx plates and to increase them in number by 

 the incorporation in the body wall of the proximal brachials and the basal seg- 

 ments of the earlier piimules, the mechanical strain caused by the long arms bemg 

 thus counteracted. 



There is a broad gap between the mechanical factors bearing upon the calyx 

 of pelagic cruioids and those influencuig the shape of the calyx of attached forms. 

 The long and supple columns of such comparatively short-armed genera as Ily- 

 crinus (fig. 3, p. 62), RJiizocrinus, Hyocrinus, Proisocrinvs (fig. 128, p. 199), Thalasso- 

 crinus (fig. 145, p. 209), and Ptilocrinus (fig. 144, p. 207) allow of a great amount of 

 swaying, so that no severe strain is ever brought to bear upon the sutures between 

 the rows of caly.x plates. Motion induced by any object hitting the crown is taken 

 up by the articulations of more or less of the upper portion of the stem, and very 

 little stress is exerted on the sutures between the calyx plates. In the pentacrinites 

 the stem, though exceedingly long, is furnished throughout with cirri, by means of 

 which the animal is attached. The result of this method of attachment is exactly 

 the same as if the stem were very short, for all the cirri which can reach a fixed 

 object cling to it, and only a small portion of the column reaches free above the 

 topmost of the clinging cirri. Thus the swaymg of the pentacrinito crown, which 

 is very large, with very long arms, is nothing like so free as the swaying of the 

 crowns of the species without cirri; the resulting added stress on the calyx plates 

 has had the effect of reducing them in size and of modifying their arrangement, so 

 that they have come to form a compact patina supporting the visceral mass and 

 serving as an attachment for the arms. In the comatulids the attachment is bj' 

 very numerous cirri, all arismg from a single plate which, mechanically, is an inte- 

 gral part of the calyx (figs. 87, p. 143, and 88, p. 145). This method of attach- 

 ment is almost as unyielding as that seen in Holopus, which possesses a stout, 

 thick, unjointed stalk (fig. 517, pi. 1); and we find, exactly as in Holopus, a maxi- 

 mum reduction of the calyx, the radials, as in Holopus, resting directly upon the 

 column, or what remains of and represents the column, the basals, as well as the 

 infrabasals, having been eliminated from the body wall altogether. 



In the gradual evolution of the perfected crinoid type (fig. 74, p. 127) the een- 

 trale was the first to become affected; fixation took place by this plate, which 

 increased in size, and became reduplicated by the continuous formation of similar 

 plates just within it, resulting in a series of columnars. 



Next the infrabasals became reduced in size, at the same time moving inward 

 toward the center over the outer border of the centraie, now become the stem (as 

 a result of the mechanical necessity of affording a firm support to the heavy calca- 

 reous body wall resting upon the now rigid reduced centraie), and gradually reclining 

 to a horizontal position, until they became merely five quite functionless minute 

 plates cappmg the ends of the basals and entirely covered by the stem, as in 



