336 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



cerated or unmutilated specimens, proving that the epithelial nerve plexus forms 

 no part of the central mechanism. Tlie I'emoval of the central organ completely 

 destroys the coordinating mechanism between the arms as tested by (1) the power 

 of executing the normal swimming movements, and (2) the power of righting 

 itself when inverted, both of these powers being permanently desti'oyed by the 

 operation. But the central capsule does not form the sole physiological con- 

 nection between the axial cords of the several arms, for experiments show a physio- 

 logical connection through the intraradial commissure. Destruction of the pen- 

 tagonal commissure as well as of the central organ destroys the physiological 

 connection between the arm pairs, but does not disturb the connection between 

 the two arms of each pair. Eviscerated individuals yield the same results as 

 unmutilated specimens. From the results of his experiments Marshall was led 

 to believe that the transverse commissure in the IBr axillary furnishes the real 

 physiological connection between the two arms of each pair, and that the chiasma 

 (the two crossed intraaxillary connectives between the axial cords of each arm 

 pair) is connected with the alternating movements of the two arms of each pair 

 which occur in the act of swimming. 



INTEKBRACHIAL Mt:SCLES. 



The interbrachial muscles of the comatulids appear on the ventral surface 

 of the arms from the radials outward as yellow or lighter or darker brown 

 bundles of fibers which form, according to their location, narrowly oblong, wedge- 

 shaped, irregularly elliptical, more or less rhombic, or semicircular masses with 

 the longer axis extending, usually more or less diagonally, quite across the arm, 

 and the shorter, parallel to the direction of the fibers, stretching between the 

 narrow ventral extensions of the brachials which form the muscle plates. In 

 species in which, as in the smaller Antedonidse, the ventral perisome is thin and 

 more or less transparent, the muscle bundles, owing to their relatively dark color, 

 are usually easily visible through it; in others, as the larger Comasteridse, in 

 which the perisome is thick and opaque, they are hidden from external view. 



The individual fibers are more or less band-like, usually elliptical in cross 

 section, and are arranged in groups with their broader sides in contact. They 

 are easily separated, and the mj'ofibrillar bundles of which each fiber is composed 

 are easily broken apart. 



Bosshard found in fresh teased preparations and in longitudinal sections 

 through the ventral arm muscle of Antedon smooth fibers, fibers with distinct 

 longitudinal striping, and fibers with double diagonal striping. 



Reich en sperger found that, after being subjected to a somewhat complicated 

 staining process, the sarcoplasma of the muscle fiber (the isotropic sub.stance of 

 Bosshard) shows a yellow or reddish coloring, and wound about it spirally are 

 the sharply outlined dark threads of contractile substance, the myofibrillw. 



The course of these spiral threads differs according to the state of contraction 

 of the muscles. If the fiber is very strongly contracted they run almost parallel 

 transversely across the muscle, though a state of such strong contraction is \evy 

 rare in Antedon and was not observed bv Bosshard. In the same muscle mass he 



