432 ROSS G. HARRISON 
five days after operation (fig. 14, taken from another specimen 
of like age). In this the shoulder elements are united into a 
single cartilage, extending from the level of the dorsal surface 
of the medullary cord to the ventral mid line. On the un- 
operated side the cartilage is considerably stouter. 
Taken together with the results of Detwiler’s studies on the 
development of the shoulder girdle after extirpation of various 
portions of the limb region, these cases indicate that the opera- 
tion used in extirpating the limb rudiment leaves in the embryo 
the rudiments of the suprascapula, the coracoid and the pro- 
coracoid. These differentiate into small cartilages, which by a 
process of hyperplasia gradually extend across the gap inter- 
vening between them until they become united into a girdle 
which is topographically complete. 
THE EFFECT OF COVERING THE WOUND WITH SKIN FROM 
ANOTHER REGION OF THE BODY 
In order to block the stimulus to regeneration, presumed to 
arise from the presence of a defect, the wound made in extirpat- 
ing the limb rudiment was in a number of cases covered with 
indifferent skin from the body of another embryo. The cover- 
ing was taken either from the flank, tailward from the anterior 
limb region, or from the belly, and consisted of ectoderm with 
at least a few mesoderm cells attached. In transplanting em- 
bryonic skin it is necessary to work quickly. The major opera- 
tion must be done first and the embryo placed in position ready 
to receive the graft. Then the skin is excised from another 
specimen, and transferred directly to the wound in the first 
embryo, where it must be pressed into place immediately by a 
silver or glass bar. If this is not done quickly the skin rolls 
itself into a ball and cannot be used. 
Such experiments were made mostly with cleaned wounds of 
various sizes. The results are given in Table 4, which should 
be compared line for line with Table 1. From the comparison 
of the two tables it is seen that covering the wound has a strong 
tendency to inhibit. regeneration. For cleaned wounds of 3 
somites in diameter the percentage of regenerating cases is re- 
duced from 52 to 33. When the wound is 33 somites or more 
