268 Bayliss and Starling^, 



In the mammalian heart the duration of the contraction and that of 

 the electrical change is much shorter than in the frog's heart, and the 

 second phase follows the first immediately, no equipotential phase being 

 apparent. It may be made to appear by cooling the heart as a whole, 

 by which means the rate of propagation is slowed and the duration 

 of the electrical change increased. It is possible that if the move- 

 ment of the meniscus in our electrometer had been sufficiently rapid, 

 we should have found an equipotential interval in all cases. With 

 our electrometer the mercury had not returned to zero before the se- 

 cond phase commenced; in some cases (see Plate XV. figs. 9 and 10) 

 with a small excursion of the meniscus, we have obtained an equipoten- 

 tial interval in the normally beating heart. 



There is another circumstance which may have some bearing on 

 this absence of an equipotential interval. The curves given by Sander- 

 son and Page were obtained by artificial excitation of the heart brought 

 to a standstill by a Stannius ligature, whereas photographs of the elec- 

 trical change of the spontaneously beating heart of frog and tortoise, 

 obtained by one of us (Bayliss) in conjunction with Prof Schäfer, 

 show no equipotential interval. (Plate XV. figs. 11, 12, and 13.) 

 Whether this was due to the slowness of the electrometer, we cannot 

 say; but it seems to suggest some difference between natural and 

 artificial excitation. 



In reference to the effects of heat and cold described above, a 

 few experiments made on the tortoise heart are of interest. Unfortu- 

 nately we had difficulty in obtaining tortoises, and were unable to re- 

 peat these experiments as we should have wished. We cut a strip of 

 ventricular muscle and laid it on the top of a warming and cooling 

 apparatus, similar to that described by Sanderson and Page, and con- 

 sisting of two brass tubes soldered together, either of which could be 

 warmed or cooled by a current of warm or iced water. The ventri- 

 cular muscle was led off in such a manner that the neighbourhood of 

 either electrode could be warmed or cooled at will, and it was exci- 

 ted by induction shocks near one electrode. We found that by cooling 

 the electrode nearest the excited point, we could reverse the electrical 

 effect, that is, whilst of course with equal temperature of the two parts, 



