INFLUENCE OF TEMPERATURE UPON LIFE-PROCESSES 427 



rapidly conducting sciatic nerves of a frog. The following table 

 summarizes the results: 



Temperature-interval . 

 -0.5to 9.5 

 0. to 10. 

 1. to 11. 

 to 13. 

 to 15. 

 to 16. 

 to 19. 

 to 21. 

 11.5to 21.5 

 12.5to 22.5 

 13. to 23. 

 14. to 24. 

 16. to 26. 



Temperature-coefficient. 

 . . 2.14 

 .79 

 .98 

 .07 

 .29 

 .57 

 .32 

 .47 

 .54 

 .67 

 .65 

 .32 

 .81 



It will be seen that almost the only coefficients approaching the 

 numerical value of 2 are those obtained at the lowest temperature- 

 ranges, 1 nor is this a peculiarity of the type of nerve-fiber employed by 

 Maxwell, for the later investigations of Lucas and Gantor on the 

 transmission of impulses in the motor-nerves of frogs bear similar 

 testimony. The following are the results obtained by Gantor: 



Temperature-coefficient for 10 interval. 

 Experimental series number. 



Average. 



1.87 

 1.77 

 1.71 

 1.64 

 1.61 



These coefficients are intermediate in value between those usually 

 obtained in physical phenomena and those which may characterize 

 chemical transformations. We are therefore led to v suspect that physi- 

 cal events play a large part in determining the rate of transmission of 

 nervous impulses. This view is rendered the more probable by the 

 historical difficulty which has been encountered in demonstrating the 

 existence of any metabolic changes in nerve-fibers or their enhance- 

 ment by stimulation, and while the recent results of Tashiro demon- 

 strate a minute evolution of carbon dioxide from excised nerves, it 

 cannot be regarded as proved that this metabolic activity is very closely 

 associated with the conduction of the stimulus. It may, rather, be 

 concerned with the maintenance of the nutrition or repair of the nerve, 

 and the inability of nerve-fibers to display fatigue on repeated stimula- 

 tion lends strong encouragement to this view, for evidently no material 



1 A small number of coefficients exceeding 2 are attributed by Maxwell to experi- 

 mental errors and are not included in the above averages. 



