E. T. Halnan 173 



which requirement would be independent of the body surface and only 

 a function of the temperature of the cell, and {'2) the eneri!;y lost throui^h 

 difference of temperature of the animal and its surroundings, and 

 depending obviously on the extent and nature of the surface. Where 

 the temperature difEereuce between the animal and its surroundings 

 is such that all loss due to the second influence is cut out, we should 

 expect the energy requirement to be proportional to the body weight 

 rather than to the surface. 



The data upon which Rubner's law was established are based 

 on calorimetry experiments on dogs whose weights varied from 3-2 to 

 31-2 kg. Leaving out the two dogs of smallest weight the results 

 obtained could be expressed equally well as a straight hne. In other 

 words, between 10 and 30 kilos., the differences between the energy 

 requirement expressed as a linear function of the body weight and the 

 energy requirement expressed as a linear function of the body surface 

 are within the limits of experimental error. This is shown much more 

 clearly when we deal with the data for oxen. As expressed in terms of 

 starch equivalent for maintenance,- the values calculated for oxen of 

 different weight by means of Rubner's surface law form such a flat 

 curve between the weights taken, that Wood and Yule found it 

 sufficiently accurate to express these results by a straight line. 

 Therefore, for oxen between the weights taken the maintenance 

 requirements are, \vithin the Umits of experimental error, a linear 

 function of the body weight, and may be expressed by the simple 

 graph given. Murray, in his paper, proposes to express this perfectly 

 simple relation between body weight and energy requirement by the 

 somewhat complicated formula 



E= flogil/- M9723 



where M = Hve weight of animal and E = maintenance starch 

 equivalent. 



This formula is obviously not based upon the graph given but 

 ujjon consideration of Rubner's law of surface area, and the close 

 relation between the results calculated from Murray's formula and 

 the figures taken from the graph emphasize the sufficiency of the 

 simple straight line graph of Wood and Yule for the expression of 

 the maintenance requirements of oxen between the weights shown. 

 As calculated by Murray, the concordance is as follows: 



M 800 1b. 1000 lb. 12001b. UOO lb. 16001b. 18001b. 



E (graph) 5-5 6-35 7-2 8-0 8-8 9-4 



E (formula) 5-47 0-35 717 7'J5 8-(i'J 9-4 



