THE HISTOLOGY OF DISSEMINATED SCLEROSIS. 595 



disseminated sclerosis. In comparison with the adjoining cortex on either side, with 

 the cortex of adjoining convolutions, or with the cortex of an altogether different 

 portion of the hemisphere, there seemed to be no change in the ganglion cells, glia 

 cells, or axis cylinders. Yet there were many areas in which recognisable alterations 

 in these structural elements in the demyelinated area did occur (figs. 299, 386-396), 

 and it is these which will now be described. It has seemed impossible to reach any 

 explanation of why certain areas showed changes and others did not do so. 



Minute areas were found in all the layers of the cortex, but before referring to 

 these it will be simpler to trace the changes in an area which passes over from the 

 subcortical white matter into the cortical layers. The earlier writers, as we have 

 seen, asserted that this transition never occurred, and that the grey matter formed a 

 barrier to the extension of the process. Sections stained with Weigert's glia method 

 were largely responsible for this statement, for such preparations showed an almost 

 abrupt cessation of the glia fibril formation at this border, and recent writers 

 emphasise the complete absence of a glial sclerosis as one of the essential character- 

 istics of a cortical area. Is, then, the process that attacks the cortex different in its 

 nature and origin from that which affects the rest of the central nervous system ? 

 Those who see in disseminated sclerosis a primary proliferation of the glia must 

 admit that in the cortex, at all events, this is not the origin of the process, and that 

 a primary degeneration of the myelin sheath is often the sole change. The examina- 

 tion, however, of a very large number of sections from very numerous areas in 

 several cases, by means of the Heidenhain iron-heematoxylin method, both in celloidin 

 and paraffin and frozen sections, and a comparison of such sections from the same 

 block of tissue, stained by Ford-Robertson's methyl-violet method, Scharlach R., 

 Bielschowsky's silver impregnation method, and the routine diffuse stains, has led 

 to the conclusion that, while a fibril formation in the layers of the cortex is propor- 

 tionate in its development to the normal glia fibril content of the layers, a change in 

 the glia cells and fine glia reticulum of the cortex is very closely related to the loss 

 of its myelin fibre content. 



In the area represented in fig. 286 it is seen that the absence of the myelin 

 affects a portion of the medullary white matter and the radiating fibres in the 

 cortex. The demyelinated tissue is roughly wedge-shaped, with its broad apex in 

 the white matter and its base on the surface of the convolution. Its outline is clearly 

 defined from the surrounding radiating fibres. The great majority of areas in which 

 changes in the cortical layers were found in association with subcortical changes were 

 in the condition of recent areas. The structure of that portion in the white matter, 

 therefore, was similar to that described under heading 2. The enormous spider cell 

 proliferation and fat granule cell formation attained, at this transition border (fig. 

 389), its maximum intensity, and passed over into the deepest layers of the cortex 

 so that the border could no longer be recognised. Specially prominent was the 

 development of the glia cell processes which attached themselves to the walls of the 



