822 
ROSS GRANVILLE HARRISON 
tion. The cell resembles strongly a spinal or cranial ganglion 
cell, and as a matter of fact this particular one was derived from 
a piece of branchial ectoderm, but the case is not specific, for simi- 
lar cells have been found in other tissue, even in pieces of muscle 
plate, and drawn out fibers of this kind have been observed in 
large numbers of cases, though more frequently in nervous tissue 
than in any other. They demonstrate the great extensibilit} 
of the embryonic protoplasm and show how a nerve fiber may be 
passively drawn out to enormous length, as no doubt occurs in 
the embryonic body, for instance, in the case of the r. lateralis vagi. 
Frequently large numbers of cells become loosened from the 
main mass and scatter themselves in the periphery. Such cells 
may remain separate or they may often be connected with one 
another by hyaline protoplasmic filaments (fig. 27). In man\ 
cases, however, the connections are more apparent than real, 
as cells that seem to have been joined will frequently glide apart 
and demonstrate the supposed continuity to be merely a close 
contact. 
In a few cases the long nerve fibers have apparently formed 
distinct nets. One of the most remarkable observed is shown 
in fig. 29. There are a large number of free endings in the group, 
as well as many apparent anastom_oses. This specimen was 
observed late one evening, and while it was under observation one 
of the connections (x) was actually resolved. On the following 
morning it was found that, with two possible exceptions, all of 
the anastomoses had been severed, each fiber being independ- 
ent of the others. Many of these apparent protoplasmic fusions 
were obviously to be accounted for by our optical limitations. 
The protoplasmic filaments are very delicate, colorless, and with- 
out visible limiting membrane. When two such structures touch 
one another, an appearance of fusion is readily given and one must 
be extremely cautious in interpreting observations. On the 
other hand, there are undoubted instances of strong adhesion be- 
tween the cells by means of filaments, as shown by the tension 
upon them if the cells move apart, but it must be borne in mind 
that very slight differences in the physical properties of the pro- 
toplasm of which the cells are composed would suffice to permit 
