PHYSIOLOGY: S. TASHIRO 
113 
pares the corresponding nerves of different animals. The data for such 
a generalization must necessarily be cumulative. The limited data we 
have secured indicate that the nerves which give off more carbon di- 
oxide in the resting state conduct the nerve impulse more quickly.* 
I may add here that conditions which influence the speed of the 
nerve impulse modify the chemical activity of the resting nerve. It 
has been known for a long time that the temperature coefificient of 
velocity of the nerve impulse is greater than that of most purely physical 
processes. We find that the temperature coefficient of carbon dioxide 
production of the non-stimulated nerve is of about the same magnitude 
as that of the velocity of the nerve impulse. 
Basing our conclusions on the foregoing experimental facts, we con- 
sider the nature of the nerve impulse as follows: 
There are two chemical conditions necessary to enable the nerve to 
conduct the nerve impulse. 
The first condition is the maintenance of normal chemical activity, 
i.e., the presence of certain chemically unstable substances. Just what 
these substances are, we do not know. This condition constitutes, to 
my mind, the so-called state of excitability. The nerve must be in 
this condition to be capable of response to a stimulus. The insta- 
bility of these substances enables the nerve to undergo greater chemical 
changes when stimulated; that is, stimulation is accompanied by an 
increase in chemical activity. This is true not only for nerve tissue, 
but also for all other living tissues, including living seeds. The second 
condition is, if this increase in chemical change at the point of stimulus 
is sufficiently greater than that in its neighborhood, the impulse will 
go in that direction. The normal nerve impulse, therefore, will go only 
in one definite direction. The possibility of nerve conduction in two 
* In this connection it is interesting to note that carbon dioxide production from medul- 
lated nerves is not always greater than from non-medullated ones, in spite of the fact that 
medullated nerves invariably carry the nerve impulse much more quickly than non medul- 
lated ones. This is rather important in view of the fact that the conducting mass in the 
medullated fiber is known to be much less than that in the non-medullated, if we compare the 
total nerve fibers gram for gram. It is very likely that if we could compare the carbon' 
dioxide production from the reacting masses of the two nerve fibers, the relative rate of me- 
tabolism of the axis cylinder of the medullated fiber would be greater than that of the non- 
meduUated. Since such a comparison is practically impossible at present, the only way to 
test the correctness of our hypothesis will be to make a series of carbon dioxide measurements 
on various medullated fibers and compare them with their corresponding rates of nerve im- 
pulse. If this relation proves to be general, then not only our current notion that the function 
of the medullary sheath is to supply nutrition to the conducting medium will be proved cor- 
rect, but also we may easily understand the morphological development of the medullary 
sheath in relation to the functional activity of the nerve fiber. 
