144 Mr. E. Billings on the Structure of 



fore they should support the bases of the ambulacra. A little 

 consideration, however, will enable any one to perceive that in 

 Pentremites the bases of the ambulacra are situated in the apex 

 of the fossil, and do not come into contact with the forked plates. 

 The apex of Pentremites is identical with the actinal centre of 

 sea-urchins and starfishes, in which the mouth is situated. It 

 is here that the ambulacra originate, and grow outward by the 

 addition of new plates to their distal extremities. There can 

 be little doubt that such was the mode of growth of the ambu- 

 lacra of the Pentreraites. The smaller extremity, therefore, of 

 their ambulacra, which is received into the forked plate, is not 

 the base, but corresponds with the apex of the ambulacrum of 

 a sea-urchin or of a starfish. It also represents the tip of the 

 arm of a Crinoid. If the forked plate is radial, then the ar- 

 rangement of the ambulacrum must be the same as that which 

 would be exhibited in a Crinoid with the upper end of the arm 

 downward and resting on the first radial, w hilst the lower end 

 would be upward, the tip being formed of the second radial. 

 From this it follows that the forked plates do not belong to 

 the radial, but to the perisomatic system. 



The five deltoid plates alternate with the forked plates, and 

 are also perisomatic. 



It is not certain that the lancet plates represent any of those 

 plates which in the Crinoidea are usually called " radials," 

 They are so arranged that if they were loosened from the walls 

 of the cup, and their smaller extremities turned upward whilst 

 their bases or larger ends retained their position, they would 

 stand in a circle around the apex, as do the arms of an ordi- 

 nary Crinoid. Their bases would alternate with the apices of 

 the deltoid plates. They would form the outside of the arms, 

 whilst the grooves and pinnulfe would be inside. Each would 

 bear on its outer or dorsal aspect two elongated sacs, the two 

 hydrospires that belong to the ambulacrum. I believe that 

 the small groove in the ambulacrum of Pentremites was occu- 

 pied by the ovarian tube only. If this be true, and if, also, 

 the lancet plates represent the radial plates of the arms of the 

 Crinoids, then the arm of Pentremites would have the respira- 

 tory portion of the ambulacral system on its dorsal, and the 

 ovarian portion on its ventral aspect. 



In the true Crinoids, both the respiratory and ovarian tubes 

 are situated in the groove in the ventral side of the arm*. In 



* Thomas Say, who was the first to recognize the Blastoidea as a 

 group distinct from the Crinoidea, also supposed the function of the 

 ambulacra to bo respirator}'. lie says, " I tliink it highly probable that 

 the branchial apparatus communicated with the surrounding lluid through 

 the pores of the ambulacra by means of filamentous processes ; these may 



