THE HISTOLOGY OF DISSEMINATED SCLEROSIS. 597 



it is sufficient to emphasise these related to the glia cell and fat granule cell develop- 

 ment, for the myelin sheath and axis cylinder changes do not differ from those 

 in other areas. In the transition from white matter to cortex the deepest layer of 

 the cortex is no longer recognisable (figs. 388-390). In the layer of the deep 

 pyramids (fig. 392) and Betz cells (fig. 387) both glia cell proliferation and fat 

 granule cell formation are still very marked : from the granular layer upwards, 

 however, the intensity of the cell reaction is much less. Nests of small round cells 

 are found around nuclear remains of atrophied ganglion cells (fig. 393), and around 

 others and around the capillaries is found a fine fat granule formation in the satellite 

 cells (fig. 396), together with a very delicate glia fibril formation, which requires the 

 highest magnification to recognise. In the marginal zone there is again a marked 

 glia cell and fibril reaction. 



If the evolution of such an area be followed, it is found that in the white matter 

 and deepest layers the process often follows the lines one would expect from the 

 presence of the large, protoplasmic, potential fibril-forming glia cells. This sclerosis 

 extends to involve the deepest layer of the cortex. Above this, in the layer of deep 

 pyramids and Betz cells, the glia cell nuclei are left as the nodal points from which 

 radiate a glia fibril formation which is insufficient to cause sclerosis ; the fibrils, 

 however, merge into a very delicate reticulum, which may be the syncytium claimed 

 by Held to form the groundwork of the cortex. In the superficial layers of the 

 cortex the long, uniform, delicate processes of the glia cells also unite with this 

 reticulum, and in the surface layer the proliferated glia cells form fibrils which extend 

 downwards — also to merge with the syncytial reticulum. 



The ganglion cell changes in relation to these areas will be taken up under the 

 heading of ganglion cells. 



When an area is confined to the cortex, the changes are, as a rule, not nearly so 

 marked, especially those in the deepest layers. The demyelination may reach from 

 the surface of the convolution to varying depths (cf. figs. 269, 273), even to the 

 border between cortex and white matter. It may lie wholly within the cortex and 

 cut through a portion of the radiation or simply affect Baillarger's stripe. These 

 areas are often in association with a markedly dilated vessel which penetrates from 

 the surface almost to Baillarger's stripe, and a number of dilated changed smaller 

 vessels lie within the area. Marchi sections of such areas show that the myelin 

 sheath is not attacked as a whole : the black staining gradually increases to involve 

 the whole myelin ring, and the axis cylinder shows by its swelling a distinct partici- 

 pation in the process. The ganglion cells in such areas show everywhere nests of 

 glia cells around them : these changes are by no means confined to the actual 

 demyelinated area, but are more marked there than elsewhere, and most 

 constant in the granular cell layer (fig. 393). The glia cells in the layer of 

 the deep pyramids show the changes represented in figs. 392, 395 — a marked 



proliferation and fibril formation, which, however, is again insufficient to cause 

 TRANS. ROY. SOC. EDIN., VOL. L, PART III (NO. 18). 84 



