28 ROBERT S. ELLIS 



A further examination of tables 11 and 12 will show that the 

 cases with ages of less than fifty years have in half of the eases a 

 greater number of cells in the right hemisphere. But if we elim- 

 inate those cases which show the right hemisphere to be notice- 

 ably below normal, the difference becomes more apparent. 

 And, lastly, if w^e observe the highest values for both hemi- 

 spheres in the tables we find that in only one case, no. 14485, are 

 the values for the left hemisphere as high as the highest values 

 for the right hemisphere ; so that although the problem is a com- 

 plicated one and the evidence is not entirely conclusive, it seems 

 probable that right-handed people start life with more Purkinje 

 cells in the right hemisphere of the cerebellum. 



With this conclusion should be correlated the conclusion stated 

 above to the effect that the right molecular layer considered in 

 relation to the internal granular layer is relatively wider than 

 the left molecular laj^er. As the molecular layer is the zone in 

 which the dendrites of the Purkinje cells ramify, the relatively 

 greater wddth of the right molecular layer is a further anatomical 

 characteristic to be correlated with the superior functional effi- 

 ciency of the right half of the body musculature. 



THE GROWTH AND DEGENERATION OF THE MYELIN SHEATHS 



Sante de Sanctis ('98) studied the development of the myelin 

 sheaths in the human cerebellum during the first three months 

 of life and found that myelination of the fibers occurs earlier 

 in the vermis than in the hemispheres. I have made no special 

 study of the matter, but it is easy, even in preparations stained 

 W'ith Delafield's hematoxylin, to see that the vermis has the fibers 

 more developed. 



Lowy ('10) made a comparative study of the disappearance 

 of the external granule cells and the development of the myelin 

 sheath in low^er mammals, and he not only confirms the state- 

 ment of de Sanctis that the vermis develops in advance of the 

 hemispheres, but he shows that the disappearance of the external 

 granule cells, the development of function, and myelination are 

 closely related. This agrees with the general results of Lui 



