14 A BICYCLE ERGOMETER WITH AN ELECTRIC BRAKE 



quickly and the temperature of the copper disk recorded by placing a 

 mercurial thermometer upon it, a rapid conduction of the heat being se- 

 cured by covering the bulb of the thermometer first with a closely fitting 

 piece of sheet lead and then with a large piece of cotton batting. As soon 

 as the mercurial thermometer reached its maximum point the temperature 

 was recorded. While possibly this method did not give the exact tem- 

 perature of the inner part of the copper disk, nevertheless it was assumed 

 that the record obtained could be considered as an index of the tem- 

 perature of the copper in the different experiments. These readings of 

 temperature were taken chiefly for comparison with the magnetic tests 

 (see Part III). 1 



Occasionally it was possible on the same day to run two calibration 

 tests, either at two different speeds or at two different intensities of field 

 magnetization, in which case, immediately after taking the temperature 

 of the copper disk, the calorimeter was again closed and the second pre- 

 liminary period run. Under these conditions it was usually not necessary 

 that the second preliminary period should be so long as the first and it 

 was possible to begin the measurements on the second basis inside of 

 half or three-quarters of an hour. 



CALIBRATION TESTS OF ERGOMETER I. 

 EARLIER CALIBRATION TESTS. 



The original ergometer (which we will designate as ergometer I) was 

 calibrated frequently in the large respiration-chamber at Middletown, 

 Connecticut, in 1903, 1904, and 1905. These calibrations were pub- 

 lished by Benedict and Carpenter in an earlier publication, 2 but they are 

 reproduced here in table 1 with some slight correction and rearrangement, 

 so that they may be readily compared with the results of the later 

 calibration tests. The shortest experiment recorded was 2 hours and 21 

 minutes in length and the longest 7 hours and 10 minutes, the strength 

 of current through the ergometer magnet varying from 0.70 ampere to 

 1.25 amperes. Of particular significance are the data given in the last 

 two columns of the table, showing the number of revolutions per minute 

 and the heat per revolution. It will be observed that in these experi- 

 ments the lowest number of revolutions studied was 71 per minute and 

 the highest 102, those for 1904 and 1905 ranging almost exclusively be- 

 tween 71 and 83 revolutions per minute. The heat per revolution ranged 

 from 0.0124 to 0.0236. A comparison of the data regarding the heat per 

 revolution with the strength of current used in the various tests shows 

 that in general the higher the magnetizing current the higher the heat 

 per revolution. 



1 The influence of the temperature of surroundings on the constancy of the ergometer 

 is discussed in Part III. 



2 Benedict and Carpenter, U. S. Department of Agriculture, Office of Expt. Stas. 

 Bui. No. 208, 1909, p. 16. 



