214 H> P> KJERSCHOW-AGERSBORG. 



a few cases I have found the madreporic plate present on the right 

 anterior nearly opposite the base of ray V., but I have never found 

 more than one in Pycnopodia hcliantlwidcs. Crozier ('21) records 

 asexual reproduction in Cose 'mast erias tcnidspina (Lamk.), its 

 model number of rays in the adult being seven, and that, together 

 with this method of autotomous multiplication, madreporites are 

 also multiplied at separate points on the disk : " an assurance that 

 portions of the body separated by autotomy will each be provided 

 with a- madreporite." He abandons the notion which he held at 

 first, that "There is a physiological basis for the development of 

 one or several madreporites according to the number of rays, i.e., 

 depending on the total water requirements of the locomotive or- 

 gans." Pycnopodia, a much larger species (even though the ray 

 lengths in the two species may be the same, Pycnopodia is larger in 

 bulk than Coscinastcrias, because it has about three times more rays 

 than the latter), and with a locomotive speed about nine times 

 greater than Coscinastcrias, has usually only one madreporite. That 

 is, it gets along with only one. Sometimes autotomy is practiced in 

 Pycnopodia (Fig. 10), but not for the purpose as recorded by 

 Crozier for Coscinastcrias. Verrill ('14) reports a multimadreporic 

 case for Pycnopodia, but he considers that to be abnormal. It seems, 

 therefore, that while the anterior end in Pycnopodia is much the 

 same as in the common starfish, the position of the madreporite on 

 the dorsal disk may vary. But it is always nearer the physiological 

 anterior end. Crozier also finds this to be true for Coscinasterias. 

 The fact that Pycnopodia is bilaterally symmetric even after meta- 

 morphosis (no one has seen the larva of Pycnopodia, but it is 

 assumed that it is bilaterally symmetric as in other echinoderms) 

 may be a factor which aids it in the development of a locomotive 

 speed greater than that of other starfish, and that it consequently 

 " selects," by means of the law of least resistance, the oldest and 

 most efficient pole as anterior end. This pole, judging from the 

 relationship which the smallest rays bear to the physiological ante- 

 rior end, is at right angles to the two interpolation zones, the zones 

 of the least locomotive efficiency during the postembryonic de- 

 velopment. Ludwig's asteroid ground-plan, applied to the ordi- 

 nary starfish, does not hold completely for the twenty-rayed star- 

 fish ; his '" vorderer Interradius " becomes the " hinterer Inter- 

 radius." The physiological anterior end corresponds more nearly 



