596 DR JAMES W. DAWSON ON 



capillaries and pre-capillary vessels (fig. 436). Each of these was surrounded by 

 rings or layers of fat granule cells, and to the outermost of these the glia cell pro- 

 cesses formed almost a radial arrangement. Each ganglion cell in the stellate layer 

 and in'the layer of the deep pyramids was in a condition of marked degeneration and 

 atrophy. Many were mere ghost-like forms with no structure ; in others only the 

 large nucleus remained, and still others were replaced by nests of small round cells 

 from five to ten in number (fig. 381). The whole tissue of these deeper layers was 

 so crowded with the two proliferated cell elements, large glia cells, and fat granule 

 cells, and with the numerous dilated small vessels, that the ganglion cells were 

 almost lost sight of. Where the area involved, as in fig. 387, the Betz cells, these 

 also were found to show all stages of degenerative change, and many of them had 

 disappeared. The changes gradually receded in intensity as the upper limit of the 

 deep pyramids (figs. 392, 395) was reached : the cells in the granular layer were 

 found to be surrounded by nests of small round cells (fig. 393), but here the glia 

 spider cell and fat granule cell formation was very limited. The fat granule cells 

 present had also a finer structure than that in the deeper layers (fig. 396). 



The intensity of the process gradually lessened as the upper layers of the cortex 

 were reached. Amongst the large pyramids were found still a large number of pro- 

 liferated glia cells, but now the processes of these were of a fine, almost uniform 

 calibre, and reached a very long distance from the small nucleus from which they 

 radiated. Marchi sections showed a very characteristic appearance in these layers. 

 All the satellite cells, which are normally round and have no visible protoplasm, were 

 found to have developed a protoplasm cell body, of various shapes, often spindle- 

 shaped and star-shaped, and to surround with their processes the ganglion cell 

 (fig. 396). Their protoplasm was studded with the minutest black granules, and 

 transition forms could be traced between those surrounding the ganglion cell and 

 rounder forms lying free in the tissue spaces or in the sheaths of the larger vessels. 

 The finest capillaries were surrounded by similar granular cells of varying shapes, 

 which had evidently arisen from the rows of small round cells which in these cortical 

 layers normally surround the capillaries. It is to these cells and to the very similar 

 round cells in the white matter that Ford-Robertson has given the name mesoglia 

 cells, and to which he attributes phagocytic properties. Of their phagocytic char- 

 acter their protoplasm, loaded with fine fat granules, is a direct proof. Here also the 

 fine uniform processes of the glia cells have a very intimate relationship to the 

 vessels, especially to the very abundant capillary plexus in this region. When we 

 reach the surface layer we find again a very abundant development of glia cells, 

 many of which are multi-nucleated and in the process of fibril formation. Diffuse 

 stains — for here specific axis cylinder stains failed — showed that numerous axis 

 cylinders could be recognised, but that very many had perished in the dense 

 zone of reaction. 



To sum up briefly the changes in a recent combined subcortical and cortical area, 



