824 
ROSS GRANVILLE HARRISON 
may form a hyaline fringe (figs. 13 and 14), and spread out con- 
siderably; pieces of the central nervous system and the primordia 
of the cranial ganglia give rise to the fiber-like structures de- 
scribed in the last section; the endoderm and notochord remain 
almost inert. 
It is, of course, needless to point out here the wide occurrence 
of protoplasmic movement in the normal development of organ- 
isms, and it will suffice to mention a few special cases which bear 
more directly upon the present problem. Within the body of the 
vertebrate embryo at the stage of development under consider- 
ation there is ample evidence that this kind of movement takes 
place. It is then that the mesenchyme is beginning to form by the 
breaking down of the epithelial mesoderm and the shifting of its 
cells to regions far removed from their source; and similarly the 
cells of the ganglion crest leave their place of origin and wander for 
a considerable distance before grouping themselves together as 
the spinal ganglia. In these cases we are dealing largely with 
the movement of single cells. A notable example of the active 
movement of masses of cells is afforded by the lateral line rudi- 
ment, which, in the course of several days, extends all the wa}^ from 
the head to the tip of the tail, as experiments show, by its own 
motile force (Harrison '03). At a later period of develop- 
ment after the first nerve trunks are laid down, there is an actual 
movement of Schwann cells along the nerve fibers, as I have 
been able to observe in the tail fin of the living tadpole ('04). In 
the same object Clark ('09) has watched the growth of lymphatics 
by sprouting at their ends. This observer has not only seen the 
actual amoeboid movement of the endothelial cells, but has 
also been able to show that the movement may be stimulated 
and directed by definite bodies, such as, extravasated red blood 
corpuscles. Within the central nervous system there is undoubted 
shifting of groups of ganglion cells, such movement having been 
taken into account by Cajal ('92, '99) in his original hypothe- 
sis of chemotaxis, and more recently by Kappers ('08) in his 
papers on neurobiotaxis, though it must be admitted that the 
evidence for active movement in the two last cases is merely in- 
ferential. In the closure of wounds we have another example 
