430 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



Each fiber bundle, and not each individual fiber, possesses a nucleus, and, 

 though the individual fibers of each bundle are more or less widely separated at 

 the ends, they are all closely packed together in the middle. 



The muscles arise from elements which are quite similar to those from which 

 the dorsal ligaments are formed. The primitive muscle cells are somewhat larger 

 than the primitive ligament cells, and their plasma appears streaked rather than 

 dotted. The nuclei possess a single prominent nucleolus. Before such cells are 

 further differentiated a mitotic division almost always occurs, preceded by a dark- 

 ening of the nucleus and a disappearance of the nucleolus. 



After this division, which takes a normal course, the two daughter cells, which 

 usually lie close together, become curiously altered. The nucleus becomes larger 

 and shows two nucleoli, and the cell plasma becomes drawn out toward the two 

 ends. Soon the cell becomes still more elongated and pointed at the ends, the 

 nucleus becomes oval, and in place of the two nuclear bodies only a single larger 

 one is found. At this stage the formation of the first fibrillae occurs. 



In the next stage the number of the fibrilla? has greatly increased, and the 

 young muscle fiber forms a twisted band. Whether this twist has any connection 

 with the subsequent spiral striping has not been determined. 



Each individual fiber terminates in a distinct dark point on the calcareous 

 ground substance, but whether it comes into actual contact with the calcareous 

 matter is not known. 



In the growing fiber a special type of granulation is visible in the sarcoplasma; 

 the nucleus is elongated and shows a small vesicle, with nuclear bodies. On the 

 insertions compact plates have developed, in which the individual fibers can not 

 be distinguished. These plates accommodate themselves to the contour of the 

 meshes in which lie the skeletal parts. From the more intensive staining of the 

 plates and of the adjacent ligamentous portion of the fibers it appears almost as 

 if the attachment was brought about by a specially secreted substance. 



Minckert has pointed out that in the ontogeny of Antedon the brachials for 

 some time after their appearance do not differ in form, and also that all are united 

 in the same way. 



Regenerating brachials show the same thing, and for some time it is impos- 

 sible to decide whether a given articulation will eventually become a muscular or 

 a nonmuscular articulation. 



Skeletal structures. The rebuilding of the skeleton commences in the earliest 

 stages and proceeds very rapidly. 



In Isocrinus the regeneration of the calcareous structures is more rapid and 

 more complete than in Antedon, and in Antedon mediterranea it is more rapid in 

 animals taken from the sea than in those kept in captivity. 



DETAILS OF TUB REGENERATION OF THE CIRRI. 



The regeneration of the cirri, total or partial, proceeds along the same lines 

 as the regeneration of the arms. 



The closure of the wound takes place much less actively and is less extensive 

 than is the case in the arms, and fewer connective tissue elements and wandering 

 cells are involved. 



