188 THE RATE OF GROWTH [ch. 



statisticians call it. Thus the following wholly empirical equations 

 give the required ratios in the case of Swedish males : 



Brain-weight (gms.) = 1487-8 — 1-94 x age, or 

 = 915-06 + 2-86 X stature. 



In the two sexes, and in different races, these empirical constants 

 will be greatly changed*; and Donaldson has further shewn that 

 correlation between brain-weight and body-weight is much closer 

 in the rat than in manf. 



Mean 2-23 206 



Brandt, a very philosophical anatomist, argued some seventy 

 years ago that the brain, being essentially a hollow structure, a 

 surface rather than a mass, ought to be equated with the surface 

 rather than the mass of the animal. This we may do by taking 

 the square-root of the brain-weight and the cube-root of the body- 

 weight; and while the ratios so obtained do not point to equality, 

 they do tend to constancy, especially if we hmit our comparison to 

 similar or related animals. Or we may vary the method, and ask 

 (as Dubois has done) to what power the brain-weight must be raised 



* Biometrika, iv, pp. 13-105, 1904. 



t H. H. Donaldson, A comparison of the white rat with man, etc., Boas Memorial 

 Volume, New York, 1906, pp. 5-26. 



