THE BRAIN 175 



body were also examined as a matter of routine in twenty- 

 seven cases. Zenker's fluid was the fixative employed. The 

 staining methods were Giemsa's as used by us for other tissues 

 and Mallory's phosphotungstic acid hematoxylin for neuroglia. 

 Giemsa's stain possesses unusual merits we believe as a rou- 

 tine stain for the central nervous system. Ganglion cells and 

 neuroglia cells, with their processes, are picked out with a 

 sharpness equal to Nissl staining, and the cytoplasmic bodies, 

 Nissl granules, etc., are shown with equal clarity. All of the 

 distinguishing features and differential staining of wandering 

 cells, lymphoid cells, plasma cells, eosinophiles, mast cells, and 

 macrophages are present. Non-medullated nerve fibers appear 

 as delicate bluish stained filaments. The sheaths of medullated 

 nerves are stained pinkish. Finally, by the employment of 

 thin paraffin sections, it is possible to trace out the ramifica- 

 tions of the small capillaries, and to study the endothelium 

 of blood vessels of all sizes. 



Lesions were found in every brain in our series. The im- 

 pressiveness of the skin rash noted in life corresponds fairly 

 well with the degree of cerebral involvement noted in the 

 histological study as Ceelen has observed. However the degree 

 of the rash which persisted post mortem was not in a general 

 way related to the extent of involvement of the brain. There is 

 also a definite correspondence between the severity of mental 

 and motor disturbances shown by the patients and the number 

 and distribution of the cerebral lesions found in the histologi- 

 cal study. These observations are shown in Table XVI. Motor 

 disturbances, such as twitching, trismus, cerea flexibilitas, we 

 associate with extensive involvement of the cerebral cortex. 

 Marked mental symptoms cannot be associated with any dis- 

 tribution of the lesions, but are associated with a general ex- 

 tensive cerebral involvement. The few cases of marked cardiac 

 disturbances seem definitely to have been associated with ex- 

 tensive lesions of the medulla and probably are in part the 

 effect of capillary hemorrhages. 



The tabulation of the distribution of the proliferative lesions 

 (Table XVI) is roughly accurate. We have not been able to 



