50 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.41. 



any point; not being confined to a syzygy as in the case of the 

 Pentacrinidae. 



Perhaps the most important question relative to the shortening of 

 the cohimn is as to whether the extremely abbreviated stems, as 

 shown in figures 1 and 3, Plate 6, were produced directly by the 

 detachment of practically the entire column, or whether the stem 

 was shortened by degrees. Doubtless the primary disruption fre- 

 quently took place well up in the proximal portion of the column, but 

 in some cases where it did not we may well conceive of secondary 

 shortening having occurred. In a highly specialized type we might 

 well look for the fracture consistently to occur in the immediate 

 proximal portion of the column. In this species, however, we 

 should expect the location of the line of separation to be largely 

 fortuitous. So far as our knowledge of the species goes, it appears in 

 a general way that the older the specimen the shorter the stem. This 

 would surely indicate a shortening of the column subsequent to the 

 prunary detachment of the organism, if found to be constantly true. 



If the abbreviation of the column has been brought about by more 

 than one disruption, such action should take place in one of two 

 ways. In the first place we could have a shortening through the 

 alternate fixation and detachment of the organism. The other 

 method would be simply the droppmg off of some of the distal col- 

 umnals. A shortening of the column by the first method seems to be 

 shown by the specimen previously referred to as being the only one 

 in which attachment has been found. The column in tliis individual 

 is of considerable size and would seem to indicate reattachment sub- 

 sequent to a period of freedom; this in turn being followed by the 

 detachment of the organism. Again, Carpenter says in regard to the 

 distal portions of some of the columns that the sutures are somewhat 

 obscured and smoothed over. This indicates perhaps not so much 

 resorption as proximity to a point of attachment. The gradual loss 

 of the major portion of the column by the successive shedding of 

 columnals is quite witlun the bounds of possibility and would unfor- 

 tunately leave no distinguislnng marks. One may not conceive, how- 

 ever, of this process taking place by gradual resorption and the loss of 

 one or two columnals at a time. The specimens indicate an abrupt 

 separation of the column, resorption playing a purely secondary and 

 minor part. 



A very interesting result of this more or less complete elimination 

 of the column is the formation of a type closely comparable to the 

 members of Group II. On Plate 6, figure 4, may be seen a form in 

 which but a single partially resorbed ossicle remains attached to the 

 proximal columnal. In figure 6 this process has been carried still 

 further, resulting in the complete loss of the column with the excep- 

 tion of the proximale. It is to be noted that all traces of the axial 



