THE HISTOLOGY OF DISSEMINATED SCLEROSIS. 617 



pia to the marginal glia zone would involve occlusion of the epi-spinal or epi-cerebral 

 spaces, an occlusion which would be extended to the peri-vascular spaces at the 

 periphery, and therefore cause a hindrance to the outflow of the lymph by these 

 channels. In a recent paper the late Dr Bruce and I indicated our agreement 

 with Nissl's contention that the adventitial lymph spaces of the blood-vessels, 

 which open into the spaces of the inner layers of the pia, are the only true lymph 

 spaces of the central nervous system. It has seemed, therefore, that this assump- 

 tion of epi-spinal, epi-cerebral, and peri-vascular spaces is an error which in- 

 validates greatly Borst's views, and causes him to overestimate the part played 

 by these adhesive changes. The fundamental importance of an increased transuda- 

 tion of toxic lymph cannot be overestimated, but the evidence of a hyperlymphosis 

 is not by any means clear except in isolated cases, and its explanation when present 

 cannot always be traced to be secondary to a closure of adventitial spaces. Borst 

 thought that together with the change in the vessel wall, which allowed increased 

 transudation, there was a closure of these adventitial lymph spaces, so that the 

 lymph, unable to get back, dilates the tissue interstices, especially the peri- 

 vascular glia and the peri-vascular space. In these dilated glia meshes the first 

 changes of the nerve fibres occurred, and later, secondary to the degeneration, 

 there was a substitutive proliferation in the glia. Such areas with dilated meshes 

 he termed " Lichtungsbezirke." 



In the course of this study attention has been directed to the fact that the 

 adventitial lymph spaces showed in the early stages only a slight degree of dilata- 

 tion, that, later, they were distended by the presence of fat granule cells, and that 

 they remained dilated and filled with cell elements of various kinds till a late stage 

 of sclerosis. In the early stages there was no question of a primary proliferation 

 such as might be induced by a toxic lymph flowing in them — the cell elements of the 

 adventitia are, firstly, the fat granule cells, secondly, these, together with proliferated 

 cells of the adventitial walls — a result of the reaction to the foreign elements in the 

 spaces — and, finally, a cell infiltration of lymphocyte-like cells. In an actual sclerotic 

 area the adventitial lymph spaces were closed in the general condensation of the 

 vessel wall, but here the relations of the lymph in the narrow interstices of the dense 

 glia tissue must be very different from those in the normal tissue. The areolar zone 

 which is sometimes found around areas of dense sclerosis may be due in part to the 

 obstruction of the lymph flow from normal tissue into sclerotic tissue. 



Very numerous peri-vascular sieve-like areas (figs. 4.48-450) were found, but these 

 were frequently related to normally myelinated tissue. The change consisted in an 

 extreme dissociation of all the fibres in the adventitia, so that the space between 

 media and the peri-vascular glious zone was occupied by a connective-tissue 

 reticulum with widened meshes which contained cell elements and frequently a 

 granular deposit. There were frequently also found indications of a lymph conges- 

 tion which extended over the whole transverse section of the cord and which found 



