346 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



Fixation b}' tho centralo results in a great strain being exerted, either bj^ the 

 constant motion of the arms or by tho motion caused by waves or by other organisms, 

 along the sutures between the centrale and the first circlet of plates, and between 

 the incHvidual plates of that circlet. This is met in such genera as Holopus (figs. 

 514, 517, pi. 1) by a sohd welding together of all the calyx plates, resulting ha a 

 solid calcareous mass with no possibiUty of motion except in the tegmen or in the 

 arms. But most commonly the strain is relieved by a combination of two proc- 

 esses, the fixed base elongating into a column with many joints, giving fle.xibihty, 

 and the plates of the lowest circlet shpping inward over the ventral (upper) surface 

 of the topmost columnal (the primitive centrale), so that they are supported by a 

 considerable i)ortion of their outer surfaces instead of by their edges only, and 

 the weak vertical suture between the centrale and the plates of the lowest circlet 

 is eliminated. The horizontal sutures, by which the plates meet end to end while 

 lying parallel to the axis of the stem, are i)erfectly capable of supporting a reason- 

 able weight by a mere thickening of the adjacent plates, and thus are not altered. 



This arrangement is satisfactory for a crinoid with comparatively short arms 

 on a semirigid column, but if the column becomes very rigid, or if tlie arms become 

 very long, it is evident that a great strain will be brought upon the sutures between 

 the plates of the lowest circlet (now horizontal or nearty so) and those of the circlet 

 just above; this is met by a change in tlie second circlet of plates by wliich they 

 become braced on the first, just as the first became braced on the topmost columnal, 

 and thus cease to form a part of the calj'x wall. Tliis has happened in the penta- 

 crinites. In the comatuhds fixation is by means of very numerous cirri all arisuig 

 from a single ossicle, which act collectively as grapphng hooks (figs. 306, 307, p. 265), 

 and is much more fkra than in the case of the pentacrinites, tho crowTis of which 

 sway at tho summit of a long, broadty spiral flexible stem. The comatuhds, there- 

 fore, must solidify the calyx stiU further to meet the conditions of life under which 

 they live, and they have done this by reducing all the calyx plates to a horizontal 

 position and welding thom sohdly together by close suture or by sjniostosis. 



Atelecrinus typically does not cling to foreign objects as do most of the coma- 

 tuhds, but rests upon the ooze on a circular disk formed by the long, nearly straight 

 cirri. It is thus not subject to any great calyx strain, and has returned its basals 

 in the condition in which we find them in the pentacrinites. 



The purely mechanical origin of the reduction of the calyx plates must be con- 

 stantly borne in miud, as it may easily be seen that a comparatively small change 

 in habit may result in an enormous change in the form and in the proportions of 

 the calyx plates which is of but minor systematic significance. An excellent 

 example of this is seen in the genus Marsupites (fig. 565, pi. 7), which superficially 

 does not in any way resemble the recent comatulids, though in reality it is very 

 closely related to them. 



With this reduction circlet by circlet of the calyx, it naturally follows that, as 

 can be seen in the young developUig Antedon, the internal organs are progressively 

 extruded more and more from the calyx, until they come to lie on and to be 

 protected by, tho lower segmejits of the postradial series (fig. 74, p. 127). The 



