RESPIRATORY EXCHANGE IN DIFFERENT ANIMALS 133 



different series show practically the same average, I -04 Cal. per kg. and 

 hour with an average individual deviation, JJL = o'l Cal. or 10 per cent. 1 

 As mentioned above (p. 60) the standard deviation in a series of deter- 

 minations on a single subject, studied by Benedict and Cathcart, 

 amounts to /z, = 5 per cent. It is rather surprising, therefore, to find 

 that the individual differences are not on the whole larger. 



Comparison between Animals of the Same Species but of 

 Different Size. The "Surface Law". 



While it had often been observed before that smaller animals had 

 per unit weight a greater respiratory exchange than larger ones, a 

 quantitative study of the influence of size upon metabolism was first 

 made by Rubner [1883] on grown dogs, weighing from 30-4 to 3-1 kg. 

 Standard conditions were not rigorously maintained, but the animals 

 were resting and the surrounding temperature kept at such a height 

 that "chemical" temperature regulations could not come into play. 

 Rubner found that, calculated per kg., the metabolism increased 

 regularly with decreasing size. When, however, the surface of the 

 animals was taken into account a practically constant metabolism per 

 square metre of surface was found for all. 2 



TABLE XXXV. METABOLISM OF DOGS. CALORIES PER SQ. M. IN TWENTY-FOUR HOURS. 



] Tigerstedt writes the average of a series of 15 determinations in the form 

 1-046 + o-ioo cal., which would imply that the mean error on the average figure 1-046 

 was o-ioo cal. A recalculation of some of his material has shown me, however, that 

 what is meant is that the standard deviation of a single determination is o-ioo cal. The 



mean error of the average works out in this case as , = 0-026, and the result should 



have been given as 1*046 + 0*026. 



2 The surface (S) of an animal is approximately proportional to the square of a linear 

 dimension (e.g. length of body) while the weight (W) is similarly proportional to the third 

 power of a linear dimension. We have, therefore, S = c WS. The constant c has been 

 worked out for a number of species. It does not vary very much even in forms of very 

 different shape. For man and also for the dog we have c = 12-3, for the rabbit 12*9, the 

 horse 9-0, the rat 9-1, and the guinea-pig 8-9 (Loewy [Op.]). 



