INTERNAL WORK. 363 
results per square centimeter of surface by the use of Meeh’s for- 
mula (p. 364) he obtained the following figures, while similar 
results are also reported on guinea-pigs, rats, and birds. 
z Carbon Dioxide 
Number of Average Live | per Square Cm. 
Experiments. Weight, Kgs. of Surface, 
Mgrs. 
Ae oriia seh Oe 24.0 2.65 
Di epuntiaseess acne 8 ch 13.5 2.60 
Hee Clages aeeOes 11.5 2.81 
Be cai eva gad 9.0 2.81 
De hailey bois 6.5 2.69 
Siesicnenae wikeetae 5.0 2.57 
Ge wensamexss 3.1 2.71 
0 ee eee 2.35 2.70 
E. Voit * has recently published an extended compilation of 
results bearing upon this point, including experiments on man, dogs, 
rabbits, swine, geese, and hens, the heat production being in most 
cases computed from the metabolism of carbon and nitrogen. The 
results when computed per square meter of surface, while they 
show not inconsiderable variations in some individual cases, never- 
theless as a whole substantially confirm the conclusion that the 
fasting metabolism is in general proportional to the surface. Still 
more recently Oppenheimer ¢ has shown that the law also holds 
good for infants. 
Causes of Variations.—In comparing experiments made upon 
different animals by different observers at different times some 
variation in the results would naturally be expected. The experi- 
ments compiled by Voit were not all made at the same temperature, 
but the range in most cases is relatively small and can hardly have 
exerted any considerable influence. Differences between the differ- 
ent animals as to their normal rate of emission of heat (thickness of 
coat, quality of skin) may perhaps have also had an effect, although 
probably a small one. 
A more important source of error seems to lie, as Voit points 
out, in the computation of the results to unit surface, what is 
actually measured, of course. being the total heat production of 
the animal. In solids which are of the same shape, that is, which 
* Zeit. f. Biol., 41, 113. + Ibid., 42, 147. 
