The ‘initial’ heat production of a non-medullated nerve (
Maia
) has been reinvestigated with more rapid recording equipment than was previously available. In a single impulse at 0° C a positive heat production was observed averaging about 9 x 10
-6
cal/g nerve: this is rapid and is probably associated with the active phase of the impulse. It is followed by a rather slower heat absorption averaging about 7 x 10
-6
cal/g nerve and lasting for about 300 ms. Previous methods were too slow to do more than record the difference between the two, the ‘net heat’, viz. about 2 x 10
-6
cal/g nerve: this is about one-third greater at 0°C than at 18° C.
Maia
nerves contain fibres from 20 to 0.3
µ
in diameter, and about half the heat is probably derived from fibres less than 3.0
µ
. The velocities of impulses in them at 0° C vary from 1.4 to 0.1 m/s, so impulses reach the recording thermojunctions throughout a long interval. Thus the observed course of the heat production is the resultant of positive and negative components in different fibres, and a substantial part of each is masked. The real positive and negative heats, therefore, are substantially greater than those observed: on the most likely estimate of velocity distribution, in a single impulse at 0° C they are about 14 x 10
-6
cal/g and — 12 x 10
-6
cal/g, respectively. Heat production, like ionic interchange, is probably proportional to fibre surface, which in 1 g of
Maia
nerve is estimated as 10
4
cm
2
. If the fibre surface is taken as 50 Å thick, the heats just calculated, if reckoned per gram of surface material, are 2.8 x 10
-3
cal and — 2.4 x 10
-3
cal, respectively. The former is about the same as the heat produced per gram in a muscle twitch. During the passage of an impulse there is known to be an interchange of Na and K ions between the axoplasm and the outside fluid. When isotonic solutions of NaCl and KCl are mixed there is a production of heat. A substantial part of the heat during an impulse may be derived from the interchange of Na and K. Another part may be associated with chemical reactions occurring in the excitable membrane during the cycle of permeability change accompanying the passage of an impulse. The negative heat production is discussed. It cannot be connected with ‘pumping back’ the Na and K ions; this is a much slower process and anyhow would probably involve a positive heat production. It may be a sign of endothermic chemical reactions, representing a first (anaerobic) stage in recovery, which occur in the surface membrane following the completion of the permeability cycle. The question is considered whether the positive and negative phases of the heat production could be due to the discharge and recharge, during the action potential, of the condenser residing in the excitable membrane. The heats so calculated are of the right order of size, but on present evidence the time relations seem to be quite wrong. The amount of K which escapes per impulse from
Maia
nerve during slow repetitive stimulation at 0° C was measured. It depends greatly on frequency of stimulation; at ‘zero frequency’ it was about 9 X 10
-8
mole/g x impulse.
The onset of contraction
