36 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.41. 



sloughing of columnals, such as is hereafter described, could have 1 

 taken place. 



Subsequent to the disruption of the crinoid from its base of attach- 

 ment increase in the length of the column takes place by the constant 

 intercalation of ossicles in the proximal portion of the column. These 

 naturally are all pentalobate in section. The first of such columnals 

 formed agree in size with the proximal portion of the primary col- 

 umn, if one may so designate that portion of the stem composed of 

 round ossicles. Subsequently, however, with the. growth of the 

 theca, there is a concomitant increase in the diameter of the columnals. 

 This results in a marked disparity in size between the distal and 

 proximal portions of the stem, and in periods of rapid growth, even 

 between juxtaposed internodal series. Such a condition may well 

 account for the structures described by Thomson and Carpenter. 

 Such an explanation is only justified upon the assumption that 

 there is during the period of intercalation of new columnals no 

 considerable increase in the diameter of those already formed. I feel 

 that we are justified in considering this to be the case. 



We may well wonder in the growth of the column whether a 

 freshly intercalated columnal, having reached the size at that time 

 attained by the stem, is ever capable thereafter of any considerable 

 increase in size. I doubt if such often be possible. Whenever we 

 have an increase in the diameter of the column which is secondary 

 its nature is immediately obvious. Usually an obliteration of sutures 

 and an appreciable irregularity of surface mark such enlargement. 

 In the case of adult Isocrinus, in the distal portion of the stem, where 

 a secondary deposition of calcareous matter apparently occurs, the 

 resultant columnals are no greater in diameter than the original 

 ossicles. According to this hypothesis crinoids showing tapering 

 columns represent normal growth without loss of distal columnals. 

 On the other hand, forms in which the column is of practically 

 uniform diameter indicate a considerable loss of stem ossicles. The 

 importance of such a conclusion, were it possible definitely to estab- 

 Ush it, is immediately obvious, inasmuch as we could then postulate 

 a detached existence for the majority of the stalked Crinoidea. 



The fact that in adult detached Pentacrinidse we do not find any 

 trace of this comparatively tenuous terminal portion of the column 

 seems to indicate that it has been lost. Were it present it could only 

 be represented as greatly enlarged by a secondary deposition of 

 stereom. It could not be other than circular in cross section, for in 

 old portions of the stem, even with pentalobate instead of round 

 columnals as a base, this is the form that the ossicles tend to assume. 

 In a few species in the distal portion of adult stems are to be found 

 columnals that are approximately circular in outline. This form, 

 which closely simulates that shown by the ossicles of the young 



