CALIBRATION TESTS 



21 



The general trend of these curves can best be compared by plotting 

 them all in one diagram, and in fig. 8 this has been done. The marked 

 uniformity in the general trend of the curves is obvious, showing that 

 with a speed between 65 and 90 revolutions the heat per revolution is 

 nearly constant. Below 60 and above 90 the heat per revolution is 

 considerably less. Since in the series of calibrations previously published 

 the speed ranged only between 60 and 90 or 100, these extreme variations 

 in the heat per revolution were not observed, and not until these later 

 tests were made did they appear. They naturally provoke discussion 

 and awaken much interest. 



.024 



.023 



.022 



.021 



.020 



.019 



.018 



.017 



.01 6 



.01 5 



.014 



.013 



.012 



30 40 50 60 70 80 90 100 110 120 130 140 150 



Fig. S. Calibration curves of ergometer I for magnetizing current of 0.7 to 1.25 amperes. 

 FRICTION TESTS WITH ERGOMETER I. 



Since in certain experiments when a man is riding the machine the 

 work of "coasting" is to be measured, a determination of the internal 

 friction of the machine is desirable. The determination is made by 

 rotating the shaft at the desired speed but without exciting the magnet. 



A great fault in the earlier tests with this instrument lay in the fact 

 that the friction tests were made under extremely adverse conditions. 

 To measure a heat production of approximately 2 calories per hour with 

 an apparatus designed to measure not less than 50 calories and as high 

 as 625 calories per hour, naturally introduced an enormous percentage 



