1902.] Photographic Records of the Response of Nerve. 211 



But f (q), whenever present, is conditioned by the position of 

 A and B. 



Consequently, if, to determine v, we compare the times of zero P.D. 

 for different values of B, or of B and A, keeping P constant, the 

 results will only be reliable if all the conductors of the bundle pass 

 under A and B in both positions, but not otherwise, because then the 

 shape of the second half of each curve will be altered, and consequently 

 the time at which the ordinate of the second half of the curve of 

 positive potential difference equals the ordinate of the first half of the 

 curve of negative potential difference, will also be altered. 



Hence, if the bundle of conductors is not regular in structure, v 

 should be determined by keeping A and B constant, and varying P, 

 the position of the exciting electrodes. 



Having found v, a comparison of the lengths p, a, and b gives the 

 true time t, and this compared with the measured time T shows whether 

 there is any latent period S. 



Influence of Temperature on Velocity of Propagation. 



Is it possible by comparison of photographs taken at different tem- 

 peratures, the position of the leads A and B and of the point of 

 excitation P remaining constant, to determine the variation with the 

 temperature of the velocity of propagation 1 



Here again there are two parts of the curve that might be selected 

 for comparison, viz., the commencement and the apex. It is necessary 

 to inquire whether / (v) is the only function likely to be changed so as 

 to affect the result. 



The commencement of the curve depends on / (v, t, p, it, q). 



Of these, / (v), the velocity of propagation, and / (w), the rate of 

 development, are the only functions that can possibly be affected by 

 change of temperature. But it is equally certain that they are both so 

 affected, the E.M.F. developing much more slowly at low temperatures. 

 (See fig. 12.) The determination therefore of the actual commencement 

 of the curve by extrapolation on the analysis is liable to an error 

 which in practice is negligible above 7° C, but increases rapidly below 

 that temperature. 



In our present apparatus, which will be fully described in another 

 paper, temperature is determined to o, 03 C, or 0° - 01 C. if necessary. 

 The importance of this precaution is evident from fig. 12. 



If the apex is taken as the point of comparison, the conditions are 

 more complex, for at the apex 



+tlf(% t, p, q) - Xlf(v, t, P , e, h . tt, q) = o. 



This will depend not only on / (v) and / (tt), but also on / (0) and 

 / (/,"), both of which may be affected by temperature. 



q 2 



