516 FREDERICK S. HAMMETT. 



exerted by the others by statistical treatment. Continuing the 

 phase of comparison dealing with the sex differences. If we take 

 as a measure of this relation the sex difference in degree of change 

 in association of the first and second order coefficients from the 

 zero order correlation between brain weight and spinal cord 

 weight (734), i.e., when first body weight (r 34 1) or body length 

 (r 34 2 ) are held constant, and then when the body weight and 

 body length are both held constant (r 3 4i 2 ), it is seen that in 

 general the reduction in value is percentagely greater in the 

 male than in the female. This indicates that the conclusion 

 drawn in the preceding paragraph is justified. 



TABLE II. 



THE (PARTIAL) CORRELATION COEFFICIENTS. 



Subscript. Male. Female. 



ri2 0.872 0.783 



rn 0.676 0.710 



ru 0.809 0.726 



ri3 0.650 0.571 



rit 0.778 0.719 



TM 0.800 0.779 



rua. . 0.773 0.654 



ns 4 0.659 0.546 



rnz 0.293 0.515 



ru 4 0.082 0.334 



H42. . 0.425 0.377 



rn 3 0.606 0.391 



7"23 1 O.I69 O.O34 



rn 4. . 0.074 0.025 



nti. .. 0.253 0.353 



"24 3 O.566 O.532 



>-341. 0.584 0.545 



0.616 0.644 



M234. . 0.655 0.573 



0.043 0.384 



f\\ 53. . 0.325 0.069 



r:a u 0.027 0.201 



*"-* 0.193 0.399 



tt4 12. . 0.567 0.570 



(i) Body Weight; (2) Body Length; (3) Brain Weight; (4) Spinal Cord Weight. 



Further confirmation of the conception is had from the fact 

 that the growth of the brain and spinal cord of the female albino 

 rat follows the changes in growth retardation in body weight and 



