836 
ROSS GRANVILLE HARRISON 
that they are marked out by the configuration of other organs. 
Grooves or spaces between the more sohd embryonic organs seem 
to be paths of predilection. Thus the dorsal nerves, j ust described, 
after leaving the medullar}^ cord, run in the small spaces left be- 
tween adj acent myotomes and the epidermis. Likewise the spinal 
nerves — at first only the motor constituents — run ventrally in 
the groove between successive mj^otomes on the inner side, to 
reach the extreme ventral part of the musculature. In these 
cases it does not seem necessary to assume that any special direc- 
tive factors of a chemotactic nature play a part. However, we are 
far from being justified in generalizing too freely from the facts 
just stated, for while the nerves which have been mentioned ap- 
parently follow paths of low mechanical resistance, others again, 
grow where the resistance is probably considerable, as when the 
first fibers in the central nervous system bore their way through 
the solid ependyma cells. 
A striking feature of the development of the peripheral nervous 
system is the fact that the principal nerve paths are laid down 
very early. In the frog the main branches of the cranial nerves, 
the sensory spinal nerves from the dorsal cells, and the motor spinal 
nerves are all formed within two or three days of the closure of 
the medullary folds, a considerable time before the complete 
absorption of the yolk. The sensory nerves from the spinal gan- 
glia follow the above named nerves after an interval of a day or 
two. The point to be emphasized in this connection is that none 
of the peripheral nerves have very great distances to grow before 
connecting with their proper end organs.-^ The . ophthalmicus 
" Harrison, '01. 
2^ In criticising a previous statement to this effect Hensen ('08) expressed him- 
self as follows: "Ich glaube mich richtig auszudriicken, wenn ich sage: es ist 
kiimmerlich sich damit helfen zu wollen dass nur kurze Wegstrecken (Wie kurz 
doch wohl?) zu durchwachsen sind. Der Zusammenhang muss zwangsmdssig 
gesichert sein." When we consider that, as the present experiments show, nerve 
fibers have power of independent growth of over a millimeter, that there are 
obviously conditions in the embryo which may direct this growth for distances of 
that magnitude, and that scarcely any peripheral nerves have normally much 
greater distances to grow, then it does not seem to me to be either futile or pre- 
posterous to assume that such conditions are sufficient to conduct the develop- 
ing fiber to its proper end organ. 
