578 DR JAMES W. DAWSON ON 



it will be necessary to refer to stages. These, it must be admitted, are somewhat 

 artificial, yet they are marked in general by definite histological characteristics. 



(1) The first indications of change (fig. 8) are best brought out by Van Gieson's 

 stain and seem to us related to the glia cells. The normal spider cells show a distinct 

 enlargement not only of their nuclei, but of their protoplasm and protoplasmic pro- 

 cesses, and the small darkly-stained glia nuclei in the tissue also show a lighter 

 staining. As yet there is no proliferation of such cells nor of any of the tissue cells, 

 e.g. cells in the blood-vessel walls. Very closely related to this glia cell enlarge- 

 ment is a change in the nerve fibres and in the glia reticulum. This seems to start 

 in a slight cedematous swelling of the tissue meshes, a swelling and faint staining 

 of the myelin sheath, and a diffuse pink staining and swelling of the contained axis 

 cylinder. These alterations are very slight, and though analogous in kind to the 

 similar changes found in acute myelitis, are not so in degree. The small capillaries 

 are dilated and engorged with blood, and show a slight participation, in the dilata- 

 tion of their adventitial lymphatic spaces, in the slight oedema of this localised area. 



This swelling of the structural elements increases till large protoplasmic glia cells 

 are formed, some of which show indications, in the presence of two nuclei, of a 

 previous mitosis, though we have never been able to recognise definite mitotic 

 figures, and it is possible, as many writers assert, that a direct division of such cells 

 may occur. The myelin sheath of individual nerve fibres is so swollen and faintly 

 staining as to be unrecognisable ; around others only a faintly-staining ring of 

 myelin can be found, and in some the swollen axis cylinder lies apparently free at 

 one side of the distended glia meshes. Under low power the nuclear content of this 

 area is already definitely increased. These nuclei belong in part to the enlarged 

 spider cells, in part to the swollen glia nuclei, and in small part to an increase in the 

 endothelial nuclei of the vessels, and, where a distinct adventitia is present, to 

 an increase in the lining cells of the adventitial spaces. The share these vessel 

 cells take is extremely difficult to decide, for numerous areas have been examined 

 in which, under low power, the affected tissue was found distinctly to contain 

 more nuclei, and yet none could be traced to any change in the cells of any 

 of the vessel walls. 



(2) The next steps in the process (figs. 9 and 349) are characterised by the 

 presence of a few large " epitheloid " cells, which, after the extraction of their 

 contents with alcohol, appear vacuolated. This stage may be termed that of a 

 commencing formation of fat granule cells. At first these are very isolated in the 

 tissue, but as increasing numbers of myelin sheaths undergo degeneration their 

 number increases very rapidly. The development of such cells at this stage must 

 be largely traced to an increase in size of the small darkly-staining glia nuclei : the 

 change in the nucleus of, and the increase in the protoplasm around, these cells may 

 be followed till round cells are found, with a central nucleus in which the chromatin 

 structure is quite visible and, with the protoplasm, taking a faint hematoxylin tinge. 



