OUTGROWTH OF THE NERVE FIBER 
797 
cannot always be made out with such clearness, owing to the exist- 
ence of the branched mesenchyme cells in their immediate vicinity. 
In other nerves, as in the case of the r. lateralis vagi of Amblystoma, 
.jthere is a slight enlargement at the end of the growing fiber, 
though branched filaments are not clearly shown there. 
It is a striking fact that in these early stages of development, 
each nerve fiber, in fact each branch of a nerve fiber, ends in an 
enlargement of this kind. The enlarged ends, as well as the fibers 
throughout their whole length, are attached to surroundingorgans 
by fine threads, but, as stated previously, I am unable to find any 
safe criterion to distinguish between natural protoplasmic fila- 
ments and products of coagulation. Aside from these fine fila- 
ments, the nerve fibers are found to end free, and anastomoses 
between different nerves are not present at this stage. This is 
perfectl}^ clear in the case of the cutaneous nerves formed by the 
cells of Rohon-Beard. A. little later, however, as seen in a R. 
pipiens embryo, 6 mm. long, the branches of the individual seg- 
mental nerves are found to have extended so far as to come into 
contact with those of the next segment, the result being the 
formation of a beautiful plexus of nerve fibers beneath the skin 
overlying the muscle plates. This is composed of fibers devoid 
of sheath cells, and in specimens hardened and stained by vom 
Rath's picro-platino-osmo-acetic mixture, the fibrillae are shown 
very clearly. Plexus formation is thus seen to be secondary, 
resulting from the accidental coming together of the growing ends 
of nerve fibers which have origin in different segmental nerves 
(text fig. 1). In all cases the nerve fibers are found to extend 
gradually out from the center, and the end of each small twig 
is characterized b}^ an enlargement made up of hyaline proto- 
plasm, provided with fine filaments, just as the main stem of the 
fiber itself is at first. 
The above observations upon the ends of the developing nerves 
agree substantially with those of Ramon y Cajal, although they 
are based upon specimens preserved by entirely different methods. 
The enlarged ending provided with protoplasmic filaments is in 
all probabihty the cone d'accroissement first described by him, the 
filaments being shown in these cases perhaps more completely 
