METHODS OF MEASUREMENT. 33 



records, which were that much higher than the kymograph records 

 (p. 32), his step-lift per minute would have been 5.0 and 5.7 meters 

 as compared with 7 meters per minute for the subject of this test. 



Considering the difference in the grade's and speeds, as well as the 

 difference in subjects, and the considerable lapse of time between the 

 results obtained in the research and in these later tests, there is more 

 approximation here than might have been expected. 



Step-lift technique during grade walking. The use of the fork on the 

 shoulders, as indicated in figure 1, during experiments on horizontal 

 walking, is, we believe, without serious criticism. After our series of 

 experiments had been completed, the results computed and tabulated, 

 and an analysis of the data was being made, a criticism arose to the 

 effect that the device pictured in figure 1 would not record the true step- 

 lift, but might include in its registration a not inconsiderable part of the 

 grade-lift. It appeared that the most logical procedure would be to 

 place the fork parallel with the belt of the mill and to run the cord from 

 the fork to the first pulley in a line at right angles to the belt of the mill, 

 continue it over the second pulley, and then down to the pointer. It 

 was argued that by this method no movement in the direction that the 

 belt was traveling could be registered, except that of a very long pendu- 

 lum, which, at this arc, namely, 110 cm. or more, would be negligible. 

 Consequently, during the preparation of this manuscript for publica- 

 tion, a series of experiments was made to test the effect of this change 

 n procedure. Since the results relate more particularly to the study of 

 the step-lift during grade walking, their discussion is deferred until the 

 results of the grade- walking experiments are considered. (See p. 243.) 



METHOD OF STEP-COUNTING. 



For securing a record of the number of steps taken by the subject, a 

 counter attached to the rear end of the treadmill was connected with 

 the ankle of the subject by means of a long, weak, spiral spring. 

 (See X, fig. 1.) This spring had sufficient tension to operate the 

 counter when the leg was thrown forward, but at the same time put no 

 restriction upon the movements of the subject as he walked. As it 

 was not possible to obtain the reading at the exact beginning of the 

 period, it was the practice to read the step-counter at the end of the 

 first minute of the period and again at the end of the eleventh. The 

 difference in the records, when multiplied by 2, gave the total number of 

 steps taken during the 10 minutes of a period 12 or 13 minutes in length. 



APPARATUS FOR DETERMINING THE PULSE-RATE. 



Benedict and Murschhauser, 1 in their report on the energy trans- 

 formations in horizontal walking, recorded a few pulse-rates with the 



Benedict and Murschhauser, Carnegie Inst. Wash. Pub. No. 231, 1915, pp. 54 and 55. 



