538 JULIA B. PLATT. 
connection of these cells with one another is at first merely by 
indifferent protoplasmic prolongations, such as connect the 
primitive neural crest cells with one another and with the 
brain. The cells on the path of the profundus develop nerve- 
fibres at about the same time as the anterior cranial ganglia. 
The remaining nerves that supply the skin are formed in part by 
fibrous prolongations from the cells of the ganglion with which 
they are connected, in part by the migration of bipolar cells 
from the ganglion, also by the distal addition of fibrous pro- 
longations from the deeper layer of the superficial ectoderm, 
and finally by the direct migration of cells from this layer into 
the nerve. These last elements contribute also to the forma- 
tion of the profundus. The difference between the formation 
of this nerve and that of the remaining sensory nerves consists 
chiefly in the fact that cells of the neural crest which take 
part in their formation are first gathered into a mass which 
we recognise as the Anlage of a ganglion, while neural crest 
cells participate in the formation of the profundus throughout 
its length. Even this difference is not absolute, for some of 
the scattered mesectoderm cells lying near the ganglion, but 
not included in its mass, contribute also to the formation of 
the sensory branches of the facialis. 
I speak of the rudiment of the ganglion as if already com- 
posed of ganglion cells, although in fact I doubt if one com- 
pletely developed ganglion cell exists in Necturus at the time 
this study closes. At present the cells of the ganglion resemble 
cells on the path of the nerve whose function it probably is to 
contribute to co-ordination merely by conduction, and not by 
modification of the impressions received. 
As the nerve-cell becomes fibrillar, the entire ectoderm of 
the sensory ridges of the head gives rise to a multitude of tiny 
fibres which cover the inner surface of the sensory area like 
fur, and appear finally to be swept together into bundles as if 
by currents of conduction. Pl. 36, fig. 24, shows the tiny 
threads that line the sensory ectoderm of the ventral hyoman- 
dibular, inner (mandibular) line as the nerve begins to form. 
Fig. 25 shows the general reticulum that connects the 
