EXCITATION AND CONDUCTION 77 
much less carbon dioxide than the distal portion (nearer 
to the retina), where the impulses normally originate. 
Table XI shows the quantitative results. 
Sensory dendrites—These results were at first 
surprising; but they became exceedingly interesting 
when we took into consideration the functional or 
developmental difference between the two nerves. 
The claw nerve of the spider crab is believed to be 
composed mainly of efferent fibers, while the optic 
nerve of Limulus is an almost purely afferent nerve. 
The direction of the normal nerve impulse in one 
of these nerves is, therefore, exactly opposite to its 
direction in the other. Developmentally speaking, 
however, the distal portion of the optic nerve corre- 
sponds to the proximal portion of the claw nerve in that 
these portions are in each case nearer the nerve cells 
from which the fibers come. Thus our results with the 
two opposing gradients may be subject to two alternative 
interpretations. Either the metabolic gradient may 
correspond to the developmental gradient, ie., all the 
portion nearer to the mother-cells may have a higher rate 
of metabolism, or it may correspond to the functional 
gradient, i.e., the nearer the portion is to the stimulus 
the higher is the carbon dioxide production. 
This question will be automatically solved if we 
study the metabolic gradient of a fiber whose functional 
direction is opposite to its developmental direction; 
e.g., an afferent nerve fiber lying peripherally to its 
nerve cells—i.e., a sensory dendrite—should be studied. 
Professor C. Judson Herrick kindly suggested that we 
use a lateral line nerve, or an accessory lateral. line 
nerve, of a fish. In the carp and the catfish both of 
