i62 Journal of Comparative Neukoi.ogy. 



the mean diameter (col. 4) of its medulla spinalis, in the cervi- 

 cal region at least, is only about one and one third times as 

 great as that of the horse. When compared with man, whose 

 central nervous system is well developed in proportion to body 

 weight, the elephant having a body weight fully fifty times that 

 of man, has a medulla spinalis, the transverse dimensions of 

 which, in the cervical region, are less than twice as great as 

 those of man. Finally, when compared with the white mouse 

 whose average body weight is about 20 grams, it will be found 

 that the elephant is 180,000 times heavier than the mouse, 

 while the mean diameter of its medulla spinalis is little more 

 than 13 times as great, or ratio of 1:180,000 as compared with 

 a ratio of 1:13. 



These comparisons do little more than emphasize the gen- 

 erally accepted truth that the smaller the mammal, the greater 

 is the proportional size of its central nervous system. 



It is inadequate and unsatisfactory certainly to compare 

 body weights with diameters of the medulla spinalis. The 

 actual weights of the central nervous system or even of the 

 medulla spinalis would of course give more satisfactory com- 

 parative relations, but unfortunately the weights of several of 

 the specimens, especially the elephant, were not obtainable, 

 and no records giving them have been found elsewhere. A 

 knowledge of the third dimension would enable one to arrive 

 at more expressive results, but the entire length of certain 

 of the specimens (elephant, horse and ox) not having 

 been obtained, comparisons involving this dimension are also 

 impossible. However, it may be noted that if the shape of the 

 body of the elephant is compared with that of the horse, man 

 and the mouse, or in fact, with that of any of the other members 

 of the series, the body of the elephant will appear proportionally 

 short. Kopsch's observations (loc. cit.) give the medulla spin- 

 alis as terminating at the level of the 1st sacral vertebra. This 

 is 10 segments before the exit from the canalis vertebralis of 

 the last spinal nerve. 



Since areas increase more rapidly than diameters, a com- 

 parison of the areas of the different transverse sections will give 



