696 Journal of Comparative Neurology and Psychology. 



numbers of fibers for the two legs, others show a great disparity, 

 the greater numbers in the latter instances being for the intact leg. 

 Noting the distribution of the nerve branches in which a considerable 

 lack of uniformity appears, we find that they carry medullated 

 nerve fibers to the muscles. The branches carrying no nerve fibers 

 to the muscles, namely the cutaneous and articular branches, show 

 a like number of medullated nerve fibers for each leg. In the sec- 

 tions of nerve branches for the muscles, localized blank spaces 

 appear with no medullated rings but the faint circular outlines of 

 nerve fibers which are no longer present, the simulacra of dead 

 nerve fibers. 



Apparently no medullated nerve fibers pass from the ventral roots 

 to the knee or ankle joints, since there is no variation for the operated 

 leg, but the supply is derived from the dorsal root nerve fibers. 

 The counts for the articular nerve fibers appear in Table V and 

 show the same number for the two legs. A marked paucity of 

 medullated nerve fibers to the joints is shown in frog E as it was 

 in frog IIB. A more detailed study of this articular innervation 

 is under way. In frog E there are eight medullated nerve fibers in 

 each leg passing to the knee, and to the ankle eight fibers in the 

 operated leg as against seven fibers in the intact leg. 



The distribution of the medullated nerve fibers to the segments 

 of the leg on the intact side of frog E corresponds roughly to that 

 found for frog IIB. One or two brief statements regarding this 

 distribution may be of interest. In the first place the numbers at 

 the various levels in the main trunks are less for frog E than for 

 frog IIB which has the same length but a greater weight. Erog E, 

 Table III, shows 6398 nerve fibers at the entrance to the thigh, 3415 

 nerve fibers at the entrance of the shank, and 1998 nerve fibers at 

 the entrance to the foot as against 7107, 4152, and 2497 at corre- 

 sponding levels for frog IIB, Table VI. It would seem as though 

 the differences at the various levels might easily come within those 

 of individual variation, which were found to be very considerable 

 in the instances of the frogs used for the first study, Dunn, 1900. 



A relation which may possibly be of more significance is that 

 shown by a comparison of the large nerve branches in the corre- 



