580 DR JAMES W. DAWSON ON 



vessel walls indicates that the cells containing the products of the disintegration of 

 the myelin are already commencing to be removed in the lymph sheaths of the vessels. 

 Probably as a result of the presence of these foreign bodies in the lymphatics there 

 is a reaction in the cell elements of the adventitia, with the production of a certain 

 number of small, round cells, with darkly-staining nuclei and a small amount of 

 protoplasm. These cells can be recognised in the adventitia together with the 

 presence of the fat granule cells, and it is at this stage that the tissue gives the 

 appearance of an inflammatory reaction in the vessel walls. Especially under low 

 power, when the significance of this nuclear accumulation is not recognised, it seems 

 that the tissue bears all the signs of an infiltrative myelitis. As the process advances, 

 the protoplasm around the nuclei increases in amount, and the cell content of the 

 adventitia can be differentiated into its various constituent elements (figs. 434-436) : 

 (l) the fat granule cell, with its vacuolated protoplasm and its nucleus, which has 

 now undergone regressive changes and appears darker, and its chromatin texture 

 denser, and later still the whole nucleus becomes crenated and fragmented ; (2) the 

 large vesicular nucleus of the proliferated endothelial cells ; (3) the darker, smaller 

 nuclei of the proliferated connective-tissue elements of the adventitia ; (4) small 

 lymphocyte-like cells whose nucleus is scarcely to be distinguished from the former. 



(4) The stage succeeding this (figs. 10 and 350) may be termed that of a com- 

 mencing fibril formation. Up till now the tissue at first glance appears as if large 

 round nucleated elements had simply distended the glia meshes and taken the place 

 of the nerve fibre ; that these were specially numerous around the vessels ; and that 

 in place of the few spider cells of the normal tissue with the delicate glia reticulum, 

 numerous large, proliferated, frequently multi-nucleated glia cells had arisen, whose 

 long-branching processes entwined between and around the fat granule cells. 



It is at this stage that there appear the first glia fibrils, as distinct from the glia 

 cell protoplasmic processes. This fibril formation is beautifully brought out by 

 Heidenhain's iron-hsematoxylin stain, and specimens stained by this method and Van 

 Gieson's method will be drawn upon for the description of the gradually increasing 

 sclerosis of this area, which has so many of the characters of an area of acute myelitis. 

 The essential feature of the rest of this evolution is the further development of the 

 fibrillar glia tissue, which takes place at the expense of the protoplasm and proto- 

 plasmic processes of the large spider cells. The lateral margins and terminations of the 

 processes are the first to develop into differentiated deeply-staining filaments, which 

 at first retain their connection with the cell body, but as the differentiation is com- 

 pleted they become independent of the process and of the cell body, but retain a 

 close relation to the cell nucleus. The nuclei seem finally to be the nodal points 

 from which the fibres radiate. The evolution of these fibrils will be later more fully 

 described, and here it is sufficient to state that these fibrils gradually assume a 

 longitudinal position parallel to the longitudinal direction of the original nerve fibre. 

 They are found gradually to increase in amount and to interlace and almost to form 



