46 M. L. Shorey 
A Zeiss binocular microscope was used in order to magnify the 
somites so that they could be counted, and their extent deter- 
mined without inverting the object. In order to get sufficient light 
for this within the egg-shell it is necessary to have either very 
bright sunlight or an arc lamp with a condenser. ‘The medullary 
tube and neural crest are very close to the myotomes at this period 
and the dorsal aorte are nearly beneath, so that it requires the 
utmost care not to injure the one or the other. A mechanical 
method of removing the somite intact would be desirable, but I 
was unable tg devise any means of doing this without injuring the 
adjacent parts, and I again resorted to electricity. “To insure the 
exact location of the needle an apparatus for holding the electric 
handles, devised at the University of Chicago, and described by 
Patterson in the Journal of Morphology for April, 1909, was used. 
Experiment 60. ‘This egg had been incubated 53 hours before 
the operation, in which I attempted to destroy the three brachial 
somites, and was preserved 50 hours after it. As my arrange- 
ments for lighting were at that time very inadequate the extent of 
the injury had to be judged by its effect: About the same areas 
were destroyed as when the wing bud itself is removed at its first 
appearance (Fig. 10) and the effect on the nervous system is 
practically the same. ‘The peripheral branching is less extensive 
on the injured side (compare 4 and B in Fig. 41) and the 
ventral horn is slightly smaller, both of which conditions obtain in 
embryos operated on at 3 days and preserved 2 days later. 
Experiment 84. In this embryo the operation was performed 
after 45 hours of incubation, and it is evident that all the muscula- 
ture of the second brachial somite was destroyed. It was pre- 
served 24 hours later. ‘There is aslight injury to the most dorsal 
portion of the spinal cord, but as the evidence that the direct 
injury was confined to this region is so unmistakable, I have not 
hesitated to regard this specimen as furnishing conclusive evidence 
concerning the behavior of motor neuroblasts in the absence of all 
muscles which they normally innervate. Moreover the ganglia 
are wanting on both sides throughout the brachial region, and as, 
at the time of operation, the ganglionic crest would lie in close con- 
tact with the dorsal portion of the spinal cord, it was doubtless 
