142 BULLETIN' 82, UNITED STATES NATIONAL MUSEUM. 



alter themselves sufficiently to meet the new conditions imposed, and this increas- 

 ing tension finally comes to exceed the tensile strength of those ligaments so that 

 the original arms are cast off at the synarthry between the first and second brachials 

 (more rarely, in the more specialized types, at the syzygy between the third and 

 fourth) and two or more new arms of a more specialized type are developed from 

 the stumps. The larval arms which are cast off, being composed for the greater part 

 of an enormous extension inward of the original calcareous cuticle, are in effect a 

 dermal structure incapable of further development of which the animal must rid 

 itself before normal growth can continue. Thus, in effect, the larval crinoid arms are 

 precisely equivalent to the calcified integument of the crustacean appendages, 

 which similarly must from tune to time be cast off to permit of the further develop- 

 ment of the animal. The casting off of the larval crinoid arms is therefore seen to 

 present a most striking similarity, as a physiological process, to the crustacean 

 moult. While normally only the multibrachiate comatulids discard their larval 

 arms, all of the comatulids discard their larval cirri. The new cirri which sup- 

 plant these, however, are not developed in the same situation, but always form 

 nearer the edge of the centrodorsal, that is, in terms of a bilaterally symmetrical 

 animal, more anteriorly. A precisely similar shifting in the position of the appen- 

 dages after a moult occurs in. many crustaceans and insects. The entire larval 

 column distal to the first stem syzygy is always discarded, both in the comatulids 

 and in the pentacrinites. Many instances of a similar rejection of larval structures 

 (as an example, the prolegs of lepidopterous larvae) among both the insects and the 

 crustaceans may be at once recalled. In many comatulids there appears to occur 

 from time to time, more or less normally, a shedding of the visceral mass. Dendy 

 has suggested that this may be an effort on the part of the animal to rid itself of 

 internal parasites; but it appears to me to find its most reasonable explanation as a 

 sort of growth moult comparable to the more or less extensive moulting of internal 

 structures which accompanies the shedding of the skin in the crustaceans and in 

 the insects. 



Orientation and the metameric divisions of the echinoderms. 



It has been commonly supposed that among the echinoderms the five radial 

 systems are primarily the five ambulacral systems, the interradial or interambu- 

 lacral systems being developed merely as space fillers. My studies on the crinoids, 

 however, have shown conclusively that, while the prolongations from the ventral 

 ring systems are fundamentally and primarily single and radial, the skeletal ele- 

 ments and the dorsal nerves are primarily and fundamentally double and inter- 

 radial, the two halves of each of the five interradial structures having moved away 

 from each other and having fused with the similar branches from the adjoining 

 interradial units with which they came in contact. The single radial derivatives 

 from the ventral systems have grown out upon supports each of which is formed 

 by the fusion of two halves of adjacent interradial processes, and is innervated by 

 one-half of each of the ad j acent interradial nerve trunks. In other words the metam- 

 eric divisions of the dorsal and the ventral part of the crinoid body alternate with 

 each other; for the primarily ventral structures the anibulacral areas each represent 



