572 DR JAMES W. DAWSON ON 



swollen bands, which represent axis cylinders (figs. 1, 2), and here and there are 

 large numbers of rows of finer and larger granules staining with the same tone as 

 these bands. These are the remains of disintegrated axis cylinders, which seem, 

 after the stage of severe swelling, to become broken up into large granule formations 

 and gradually into finer granules before they ultimately disappear (figs. 425, 426). 

 These remains of axis cylinders can be distinctly recognised not only with eosin, 

 picro-fuchsin, and iron-hsematoxylin, but also with the elective axis cylinder stains, 

 e.g. Cajal's and Bielschowsky's methods. 



(b) In the Cerebral White Matter. 



(i) Nerve fibres cut longitudinally (figs. 284, 304, 370, 373-376), e.g. an 

 oval area at the base of or within a medullary ray. 



Such an area is again very similar in structure to that described in the spinal 

 cord. Weigert sections show that there is a gradual transition on all sides into 

 healthy tissue (fig. 284), and that the adjoining nerve fibres show marked traces of 

 swelling, diffuse staining, and granular disintegration. The blood-vessels appear very 

 numerous and even the capillaries are dilated, engorged with blood, and, in nuclear 

 staining, show an evident increase of the endothelial nuclei, together with a very 

 abundant presence of fat granular cells in their adventitial spaces. The two most 

 characteristic features of the cord area are here again present : ( 1 ) the rows of fat 

 granule cells, which seem to occupy the tubular spaces of the degenerated myelin 

 fibres (fig. 373) ; and (2) the presence of rows of large, protoplasmic glia cells 

 (fig. 375), whose processes entwine between the rows of fat granule cells and separate 

 them into more definite layers. These processes divide into numerous, fine, inter- 

 lacing and anastomosing branches which tend to take a longitudinal direction, the 

 protoplasm of some of the cells being drawn out also in a longitudinal direction and 

 giving off numerous processes from either pole. The glia nuclei are vesicular and 

 elongated and frequently one or more are at either pole of a cell. Together with 

 the larger nuclei are found many smaller, darker-stained nuclei, with little protoplasm 

 and no processes. 



The adventitial wall of the longitudinally-running vessels is again infiltrated with 

 fat granule cells, which form a single or double layer, and when the capillaries join 

 the smaller vessels the whole tissue in the angles between the vessels seems permeated 

 with such cells. In the walls of the vessels are found other nucleated elements 

 which are rounder, denser, and have only a small cell body, and similar cells are 

 found frequently in small groups or isolated in the tissue immediately adjoining the 

 capillaries. The axis cylinders are here very thickened and vesicular and in many 

 parts transformed into clumps of larger and smaller globules and granules. In all 

 cases they seem pushed aside by the large glia cells and rows of fat granule cells, 

 and take the direction between these rows. Marchi-stained preparations (figs. 304- 

 306) bring out very beautifully the enormous number of fat granule cells and the 



