804 
ROSS GRANVILLE HARRISON 
of the previous experiments upon the development of the nervous 
system. The medullary folds are just completely closed and the 
tail bud is barely visible. The reason for choosing this stage is 
because it is the latest in which there is no histological differen- 
tiation in the nervous system or muscle plates. All of the cells 
are compact and no fibers whatever are present. The tissues are 
thus got into the lymph before their histogenetic development 
has begun. 
The transference of the tissue to the lymph drop cannot be 
accomplished without a considerable amount of tearing. Often 
single cells or small groups are torn loose from the main mass and 
individual cells are fragmented, setting free yolk and pigment 
granules, but the fibrin holds the main masses together, unless the 
lymph is too thin, in which case the embryonic cells round off and 
separate from one another. This same kind of disintegration has 
been observed also in some cases in which the clot was firm . Even 
in the absence of bacteria the cells in these specimens may remain 
entirely unchanged, manifesting none of the peculiar protoplasmic 
activities seen in successful preparations. It has not been pos- 
sible to assign any particular cause for this condition, and it must 
be attributed to slight deviations from the normal in the composi- 
tion of the medium. All such experiments, and these have formed 
but a small percentage of the whole, have been rejected as incon- 
clusive, and have been so indicated in the tabulation of results. 
While the methods of preparation were practically the same 
in all cases, the experiments themselves were varied considerably 
as regards the tissues isolated. The chief object of the work being 
to test the power of embryonic nerve cells to form fibers by out- 
growth, the largest number of experiments were made with ner- 
vous tissue. In some cases the medullary cord was dissected out 
entire, though it usually broke when transferred, and in others 
it was purposely fragmented by teasing. In quite a number of 
cases portions of the muscle plates were left attached to the med- 
ullary cord. In other experiments pieces of ectoderm from the 
branchial region, together with the underlying ganglia were taken. 
The behavior of this tissue, as regards the formation of fibers, 
was altogether similar to that of the m.edullary tube, and has 
