122 PROCEEDINGS OF THE NATIONAL MUSEUM. tol. 41. 



bringing about conditions as we know tbem, or in 'preventing any 

 considerable increase in the length of the columns, seems a reasonable 

 supposition, however. 



Batlier (1900) believes that the assumption of a detached existence 

 by the Pelmatozoa results in a marked abbreviation of the column, 

 drawing this inference from the facts as we know them in the case of 

 Millericrinus prattii. It is not evident that this case may be assumed 

 to be typical. An eleutherozoic habit in the majority of cases prob- 

 ably did not result in a marked shortening of the column. Indeed, 

 the longest columns reported are those of the fossil Pentacrinus 

 which occurred in types which probably were not attached. The 

 columns of crinoids have at no time been of great length, and it may 

 well be that the widespread maintenance of an eleutherozoic habit as 

 here assumed may have had much to do with keeping the colunms 

 within moderate bounds. It is probable only in special cases where 

 other factors enter into the question that we have an appreciable 

 abbreviation of the stems. 



A fairly long column would be of considerable advantage to a 

 crinoid whether free or attached. Of course in the case of complete 

 separation of the crown from the column these advantages are more 

 than compensated for by the additional degree of freedom. In the 

 case of stalked forms however, a moderately long stem would serve as 

 ballast either with or without a terminal weight and give the crinoid 

 a certain degree of stability while at rest. The advantage of this is 

 at once apparent. It is apparently necessary for crinoids at times to 

 be stationary. The instances of comatulids dying in a bowl of water 

 wherein there were no pebbles or other objects to which they might 

 attach themselves is well known. Other specimens similarly placed 

 but with objects to which they might attach placed in the bottom of 

 the bowl live for some time. Again, moderately long columns would 

 be of advantage as keeping the crinoids well above the sea bottom. 

 The advantages of this are obvious both to the animals themselves 

 and to the young of the colony. The impossibility of the young 

 finding a foothold in a colony of well grown Comatulae has well been 

 stated by Clark (1909, a). 



The formation of prehensive cirri may directly be attributed to 

 the maintenance of an eleutherozoic existence, however. Wherever 

 found, there we may postulate free-living crinoids. It is evident 

 that in the case of purely statozoic forms nothing could be more 

 useless than appendages designed for holding to extraneous objects. 

 In earlier times the cutI were comparatively simple, but so specialized 

 have they since become that isolated cirri of certain Pentacrinidse 

 and comatulids can be distinguished with difficulty. Even the 

 terminal claw has been developed in the stalked forms (PI. 4, fig. 2). 

 The function of these cirri has been primarily no doubt that of 



