372 PROCEEDINGS OF THE ACADEMY OF [1890. 



union of the ambulacral and interanibulacral plates, and the closure 

 of mouth and food grooves, the conditions of the two structures 

 would be almost identical. 



From these observations we conclude that the heavy plated, ex- 

 travagantly developed integument, the so-called vault, was gradually 

 evolved in geological times from the thinly plated disk of earlier 

 forms. This may have been accomplished in the following way : 

 At first the ambulacra were exposed at the surface, but subsequently 

 became covered over by the great development of interambulacral 

 plates, which encroached upon their space at the surface from each 

 side, and finally closed in above them, crowding the ambulacral 

 skeleton inward. In species in which the ambulacra are not 

 covered but remain external, the covering pieces were stronger and 

 oftered greater resistance, so that instead of being crowded inward 

 they became incorporated into the test. 



Dr. Carpenter has suggested to us in a note a somewhat similar 

 case — though he thinks the parallel must not be carried too far — in • 

 the gradual obliteration of the ambulacral grooves in some arms of 

 Adinometra, where the convex perisome gradually encroaches more 

 and more on the sides of the groove, bringing its edges together, 

 and finally closing it. It seems to us that this is a very significant il- 

 lustration of what may have happened in the evolution of the closed 

 " vault " from the open disk. 



Returning again to the Inadunata, we find in the Fistulata the 

 palaeontoloffical development essentially different from that of the 

 preceding group. The plates of the ventral side of the calyx at no 

 time attain the I'igidity and large size of those in the Camerata. 

 The plates of the dorsal side, with the exception of the anal plates, 

 which we consider afterwards se])arately, undergo scarcely any 

 changes, and the brachials in all of them are free plates from the 

 radials up. The ventral sac in some of the earliest forms is quite 

 small, but rapidly attains enormous dimensions, constituting the 

 greater part of the calyx ; but at the end of the Carboniferous period 

 dwindled down almost as rapidly again to its former insignificance, 

 so as to be representediin Cromyocrinus, Eupaehycrinus, Erisocrinus, 

 and Encrimis only by a short cone. Respiration was effected direct- 

 ly through the test, but apparently only at the posterior side of the 

 calyx, either by means of pores along the ventral sac, or by a mad- 

 reporite placed anteriorly to the sac. The ambulacra of all Fistu- 

 lata, so far as known, are tegminal ; they are bordered by side pieces, 



