192 



BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



At the time the uiternal organs begin to exhibit this excess of growth over the 

 external skeletal system the basals have more or less ceased develoj)iiig, and have 

 leaned so far outward that the mechanical stress of this excess growth falls entirely 

 upon the radials. 



In the comatulids the radials are greatly reduced, and the gradual cessation of 

 their dcvelojimcnt begins not long after the same thing has commenced to become 



evident m the basals. Thus in Pro- 

 machocrinus the radials arc unable, 

 through incipient cessation of develop- 

 ment, to grow laterally and to occupy 

 the vacant spaces left by the spreading 

 outward of the radial circlet as a result 

 of the excess of growth of the visceral 

 mass; but these spaces, exposing peri- 

 some belonging to the skeleton forming 

 dorsal body wall, become at once occu- 

 pied by narrow plates, which ra])idly 

 increase in width as the spreading apart 

 of the radials progresses. 



The water vascular system is ]iri- 

 marily a ventral system; it is thus pre- 

 pared to send an extension at once into, 

 or to be drawn out with, any process 

 arising from the dorsoventral margin. 

 Along with the water tubes the am- 

 bulacral grooves and the subambulacral 

 nerves always take advantage of any 

 extension of the perisomic surface and 

 at once extend themselves over it. 

 Evidence of this is seen at all j)oints 

 where the arms branch. 



It is therefore to be expected that 

 if the skeleton forming dorsal surface 

 of the animal gives rise to interradial 

 processes resembling the radial ]iroc- 

 esses, the ventral structures will make 

 exactly the same use of them that they 

 did originally of the radial processes. 

 It might be expected that the ambulacral systems would extend themselves 

 upon the ventral surface of the interradial arms by forming interradial buds, as 

 they do in the case of their radial extensions. But they do not d<> this. Dorsally 

 the five supernumerary radials and post-radial series of Promaclwcrinus and Thau- 

 matocrinus are truly interradial so far as the skeleton is concerned. \"entrally 

 each of the post-radial series derives its ambulacral structures not from the center 

 of the interradial portion of the cu'cumcesophageal structures opposite it, but from 



Fio. 123.— Lateral view of a specimen of Ateleceihus 



BUICATUS FROM TUE TniLIPPINE ISLANDS, SHOWING THE 

 GREATLY REDUCED BASALS, AND TUE FURROWS DELIMITING 

 THE RADIAL AREAS OF THE CENTRODOESAL. 



