258 RITTER AND CROCKER 



potentially double. Now, when a new ray is formed the whole 

 of the outer canal at the point of origin, and not its outer wall 

 merely, loops into the new ray. The two limbs of the loop 

 come in contact, and thus the two-layered septum of the new 

 radial canal is produced. The inner canal, then, has nothing 

 to do about the production of the radial canal. 



It will be readily comprehended that the growing radial water 

 canal, together with its outpocketing tube feet, is wholly em- 

 bedded in mesenchymatous tissue from which the muscles of 

 the feet, and also the ambulacral skeleton of the new ray, are 

 formed. It will also be easily understood how the coelomic epi- 

 thelium becomes closely applied to the new ampullar within the 

 ray; and the ectoderm to the tube feet on the outside. 



The early time (long before they become functional) at which 

 the tube feet take on their distinctive character, is worthy of 

 note. As already mentioned, the muscle fibers begin to appear 

 when only two pairs are blocked out ; and before they become 

 visible externally at all their terminal discs are formed. 



Although we have not attempted to trace in full the develop- 

 ment of the skeletal parts of the new rays, a few observations 

 with regard to the relation of the ray at its very outset to the 

 older parts of the oral skeleton, may be recorded. No evi- 

 dence of the resorption of any of the old skeleton, or, in fact, 

 any other tissue, has been seen. The new forming parts pen- 

 etrate rather into the fibrous septal tissue between the ad ambu- 

 lacral plates and wedge these latter apart ; and it is in and from 

 this fibrous tissue that the basal skeletal plates of the new ray 

 are formed. These appear very early, being recognizable 

 when two pairs of ambulacra are established. Not only the 

 adambulacral, but also the interbrachial ossicles show, in frontal 

 sections, evidence of this wedging apart. 



The gastric, or hepatic, diverticula for the new rays do not 

 develop early : not until the ray has reached the edge of the 

 disc at least. 



A vertical interradial section through an interradius in which 

 no ray is developing shows that while the points at which 

 growth will take place to produce the several parts of the new 

 ray are not far separated from one another, they are yet by no 



