206 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



attached by this soft integument which then lengthened out into a slender stalk, 

 resembling the stalk of such forms as Boltenia, carrying with it, of course, the spi- 

 cular investment, the included calcareous deposits of whicli increased in number 

 and became segregated into definite ossicles. Such an origin for the column would 

 accord with what we know of the origin of the cirri and of the distal portion of 

 the pinnules. 



This would make it clear at once how it is that the prolongations from the 

 chambered organ and the associated stractures became continued into the column; 

 but while there is evidence tliat something of the kind may have occurred in certain 

 of the older fossils, it does not seem to luive occurred in any of the recent types 

 nor in any of ihcir immediate fossil representatives. 



The elongation of the apical jilate as presupposed in the two first alternatives 

 does not necessarily call for a uniform deposit of stereom all over its internal surface. 

 The chambered oi^an and the accessory structures probabl}' retained their original 

 relationship with its center and became drawn out into a complex axial cord as a 

 result of the deposit of stereom about the periphery; or the new columnals, formed 

 just \vitlun the apical plate, arose as rings (as the topmost colunuials do in all of 

 the recent forms) wliich grew inward until the distal portion of the elongated 

 chambered organ was reduced to a very small diameter. 



As described above, these three possible origins of the column and of the indi- 

 vidual columnals would appear to be very different, but upon consideration it be- 

 comes evident that the difference is more in words than in fact. We are pro1)ably 

 nearest the truth if we consider that all three alternatives play a part in the for- 

 mation of the crinoid column, but place tlie greatest emphasis upon the second. 



The columns of the later and recent crinoids in general difi'er from those of the 

 earUer forms in developing with much greater rapidity, though this is masked by 

 tlio fact that they possess also a definite growth limit at the attainment of wliich 

 further development ceases, such a growth^ hmit being unlcnowoi in the palaeozoic 

 types. 



A series of loose sutures is mechanically available only for slowly growing 

 columns, in wliich the inch%adual columnals are very short. With increasing j)ro- 

 portionate length the loose sutures between the columnals gradually undergo a 

 chfl'erentiation; a fulcra! ridge develops, and the ligament fibers become segregated 

 into two large bundles, one on cither side of it. 



It is by this process that a column formed according to the second hypothesis 

 becomes transformed into the type characteristic of the later fossil and the recent 

 crinoids. 



There is a definite limit to the possibihties of further growth in a column com- 

 posed of long ossicles fastened end to end by alternating articulations consisting 

 of two ligament masses separated by a fulcral ridge. If the animal remains small 

 with a small hght crown, such a colunm may safely attain a length of 100 or more 

 columnals, but if the crown should become of large size and heavy, a stem of tliis 

 tyjjc would not be able to support it; tlie rajiidly increasing tendency to "buckle" 

 would limit the available length of a stem of this nature. 



