CRINOIDEA. 291 



characters are as follows. (1) They are all, with the partial 

 exception of the Comatulids and possibly one or two others, 

 attached throughout life, and the oral pole which also carries 

 the anus is turned upwards. (2) The gonads are removed 

 altogether from the disc and lie in the ultimate branches of 

 the arms. (3) The anterior coelom becomes merged in the 

 general perivisceral cavity, and there is no axial sinus in the 

 adult. (4) The anterior coelom is given off from the enteron 

 separately from the posterior. (5) The absence of any trace of 

 a right hydrocoel. (6) The fact that the oral surface of the adult 

 is derived from the posterior surface of the larva. (7) The form 

 of the larva. (8) The open condition of the ambulacral grooves. 

 Some of these are absolutely distinctive, viz. (4), (6), (7). Some 

 of them are shared either wholly or partly by Asteroids, viz. 

 (1), (2), (8). Holothurians present (3), and so far as is known 

 (5). The bias is therefore on the whole towards Asteroids, as 

 we have already pointed out in discussing the affinities of Holo- 

 thurians, but the bias is very slight, for (8) cannot be regarded 

 as an important character seeing that Ophiuroids, so closely 

 related to Asteroids, do not share it, and (2) is partly shared by 

 Echinoids, so far as the pentamerous arrangement of the gonads 

 is concerned. So that (1) only is left. But this resemblance 

 carries with it an important difference. It is true that Asteroids 

 are the only living Echinoderms outside the Crinoids which pre- 

 sent fixation at any time of life and that the fixation is effected 

 by the preoral lobe, but as shown by MacBride there is this im- 

 portant difference : whereas in Asteroids the pedicle of attach- 

 ment is found to arise from the oral surface of the adult and is 

 surrounded by the hydrocoel ring (Fig. 105), in Crinoids it springs 

 from the aboral surface and is outside and far removed from 

 the circumoral water- vessel. This discrepancy undoubtedly 

 receives its explanation by a consideration of the fact that fix- 

 ation takes place in both classes some time before the hydrocoel 

 ring becomes complete. But it is none the less significant, 

 especially when taken in conjunction with another fact. In all 

 Echinoderms the mouth shifts during the development. In all 

 classes except Crinoids it shifts on to the left side, so that the 

 left side or left ventral side of the larva becomes the ventral 

 side of the adult. In Crinoids however it shifts further ; not 

 only does it move on to the left side indenting the left hydrocoel 



