Mans Mechanical Efficiency in Work Performance. 407 

 therefore T = W 1>275 /9-284 = 1:17/*, 



therefore fi = W 1 ' 878 / 10*86. (k) 



But 



^ = e 6-i87/p6-its4 = W^^/IO-SG therefore P c * = 4-042/W° ,20 s. (I) 



Therefore the mE' factor and P have both been expressed in terms of the 

 " stripped weight." However, since the mass in motion includes the clothes, 

 it seems essential at this point to introduce the " clothed weight " ; and an 

 interesting way of doing so is to determine the values of P for Kemp and 

 Armstrong, from the general expression (I) given above ; and then convert 

 the values so found into terms of the clothed weight. Substituting 624 and 

 43 - 7 for W in (I), it follows that in Kemp's case P = 1-714, and in 

 Armstrong's case P = T846. Now their clothed weights (for weight of 

 clothes, see p. 397) were 634 and 47'5 kgrm. respectively ; terming this weight 

 W 2 , in both cases the given values of P e may be expressed as 



P c = 4-88/W 2 °- 252 ; t» 



substituting this new value for P c in the " (mE') " factor, 



C 6-187/ P 6-134 = (W2/9-82) 1 - 546 = (Wa/^) 1 " 546 , 



therefore mE' = m * 546 x m. (n) 



Eecognising now with some certainty the fact that this is indeed a " mass 

 factor," even if qualified by something in the nature of E', it is safe to 

 conclude that either 1/P or 1/P 2 is a unit of length. The conclusion may be 

 said to follow at once from the general cubical nature of the factor, which 

 may, indeed, be described as (e 2 /P 2 ) 3 qualified by a correction for clothing 

 (difference between e 6 * 187 and e 6-134 ) and a correction for density (difference 

 between 6434 and 6 - 000). Having arrived at this conclusion, however, it is 

 reasonable to be prepared for several different relationships to the whole mass 

 of the body resulting from temporary alterations in this individual unit of 

 length, not only in reference to the individual mass, but also to the degree of 

 shortening or extension of the body mass associated with individual move- 

 ments. In short, it is reasonable to infer that the comparison between the 

 value of P found for Briscoe and that found for Douglas is not assignable 

 merely to different body weights, but also to an essential distinction between 

 P e and P w , the " cycling " and " walking " values of P respectively. Taking 

 this view, I have considered it not unwise to assume that whereas P c is related 



* P„ denotes the " cycling value " of P. 



