264 Bayliss and Starling, 



of the capillary curves. For if the wave starts at any point A 

 whatever its time-course, it is plain that negativity of A will always 

 be the first phase on the photographic curve. Yet we found above 

 that the first wave shewing basal negativity disappeared entirely with 

 further cooling of the base. Hence, if we trust the record of our in- 



Fig. 4. 



strument, we are driven to the conclusion that in the mammalian ven- 

 tricle, the excitatory wave is distinct from the wave of negativity, 

 and may precede it, and that the excitatory wave travels from the 

 auricle to the base, and thence to the apex of the ventricle, so caus- 

 ing under normal circumstances an electrical variation beginning at 

 the base and ending at the apex. Cooling the base would both leng- 





•'«r 





Fig. 5. 



then the latent period of the variation and also alter the character 

 of its curve, so that on the electrometer the first ejßfect of cooling the 

 base would be to produce a triphasic variation, and the final effect 

 to reverse the diphasic curve altogether. 



We may conceive three ways in which tliis wave of excitation 

 might travel from base to apex: 



