156 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



After all possible compensation has been made for the elongation of the digestive 

 tube beneath the enormously enlarged anal area, furtlier ])ressure forces accommo- 

 dation bv a lateral migration of the mouth to the right, resulting in the secondary 

 bilateral symmetry. 



This ijiterpretation of the conditions found in the Comasteridje gives us a clue 

 to (he siiinificaiK-e of the anal structures characteristic of the species exliibitijig so- 

 called secondary bilateral symmetry. In these species, so far as we kjiow, the diges- 

 tive tube makes a little more than one complete turn, to the right, as in the Comas- 

 teridae; the mouth is cejitral or very nearly so, while the anal proboscis, situated in 

 an interambulacral area -which is more or less enlarged, is marginal or submarguial. 

 The digestive tube runs about the margin of the disk, its anterior portion turning 

 abruptly centralward to the mouth; this anterior portion is narrow, of more or 

 less hxed diameter, and of more or less fixed position. The middle and posterior 

 portions of the digestive tube are larger, more variable, and less fixed. Thus any 

 lengthening of the digestive tube, or any gorging with food, has the effect of alter- 

 ing the relationshi{)s of the posterior end, the anal proboscis and the surrounding 

 structures. 



In the echinoderms with a rigid covering, the echinoids, asteroids and ophiu- 

 roids, each end of the digestive tube is more or less firmly fixed; hence the accom- 

 modation necessary as a result of the motion constantly taking place is taken up 

 along its central portion within the ample body ca^^ty. In most of the holo- 

 thurians the elastic and pliable body wall admits of accommodation to internal 

 changes, wMle in the others accommodation is effected as in the urchins. In the 

 echinoids, asteroids, ophiuroids and holothurians, therefore, there is no incentive 

 to external change from the constant changes taking place in the digestive tube in 

 the exorcise of its functions. 



In the crmoids conditions are quite otherwise; here the body cavitj^ is reduced 

 to a minimum ; the dorsal part of the visceral mass is inclosed by a rigid cup and the 

 ventral part is roofed over by a pliant, though more or less plated or at least spicu- 

 liferous, tegmen. Owing to the small size of the body cavity aU the internal organs 

 wliich are unable to migrate out along the radial extensions are greatly crowded. 

 Any internal movements must therefore be accommodated by changes in the ven- 

 tral covering which, if extensive, may be communicated to the calj-x plates about 

 its border. 



I have remarked that the interambulacral area in which the anal proboscis hes 

 is always the largest of the five mterambulacral areas; its surface is also always the 

 most convex. The constant movements of the posterior end of the digestive tube 

 appear to be amply sufficient to explain this. 



Now the posterior portion of the digestive tube enters the region under the 

 posterior interambulacral area from the right; hence the tendenc}^ of the motions 

 here and of the lengthening of the digestive tube would be to shove the anal pro- 

 boscis constantly toward the loft, and also, as the digestive tube rises into the 

 anal proboscis, to pull the surface of the outer right hand side of the posterior 

 interambulacral area upward. 



