CORRELATION OF BRAIN AND SPINAL CORD. 521 



problem of differential development, there emerges a unanimity 

 of inter-relationship which substantiates the premises on which 

 the interpretations are based. 



The variability in brain weight is greater in man than in the 

 rat. This is probably another expression of the difference be- 

 tween heterogeneous and homogeneous material. It is, there- 

 fore, not essentially a real difference. Contributive to, it may be 

 the greater variability in human body weight as compared with 

 rat body weight, and the high correlation between brain weight 

 and body weight in the rat. The low association between this 

 pair of variables reported for man rather negatives this idea, 

 however. 



Both brain and spinal cord are more variable than body length. 

 In man also brain weight is more variable than stature. This 

 may be taken as an indication that skeletal composition is more 

 fixed and metabolically stable than is that of the central nervous 

 system. Such a conclusion is in accord with the chemical and 

 physiological data so far available. 



The brain appears to be less variable than the spinal cord. 

 This is consistent with the fact that brain weight is less highly 

 correlated with body w r eight of high variability than is the spinal 

 cord. On the other hand when body length variability is elimi- 

 nated from brain and spinal cord variability free from the 

 assumed influence of body weight variability, during the compu- 

 tation of the actual or reduced variability of these organs ac- 

 cording to the usual statistical procedure, there is no reduction in 

 value (of brain and spinal cord variability) below that which 

 obtains when the variability has been stabilized for body weight. 

 A different result would be expected if the respective brain 'and 

 spinal cord associations with body weight and body length were 

 dominant factors in the variability coefficients. Moreover, since 

 the spinal cord has a relatively higher percentage of the stable 

 (non-readily utilizable) lipoids characteristic of the central 

 nervous system, and a relatively lower percentage of the labile 

 readily utilizable elements than the brain, it would be expected 

 that its (the spinal cord) variability would be less than that of 

 the brain, if this compositional difference is the factor of im- 

 portance. All the evidence so far accumulated gives an afnrma- 



