LYMPHATIC AND VASCULAR RELATIONS 
31 
tinuous endothelium-lined channels. The so-called venous sinuses, i. e., 
the subcapsular or peripheral blood space as well as the central or in- 
ternal blood spaces, are, on the contrary, not bounded by a definite wall, 
and their contents are directly continuous with the parenchyma of the 
rest of the node, v Schumacher too called attention to this fact, and be- 
lieves that the wall of the sub-capsular blood space, which he regards 
as having been a true marginal lymph sinus developmentally, was proba- 
bly incomplete from the beginning. Consequently then, these fluctuating 
vascular areas could be regarded as spaces resulting wholly or in part 
from a depletion of the parenchyma of the node, with subsequent or, 
perhaps better, simultaneous substitution of blood cells. Hence these 
areas can continue to exist only as long as the withdrawal of lymphocytes 
from the node into the circulating blood exceeds their proliferation 
within the node. Moreover, if the existence of a definite endothelium 
continuous with that of the lacunae would have to be assumed for the 
central and peripheral blood spaces, then it would be difficult indeed to 
explain why the blood corpuscles are frequently distributed quite uni- 
formly among the lymphocytes throughout the whole node, and how 
such endothelial walls could so rapidly accommodate themselves to the 
comparatively rapid and very extensive fluctuations in size which they 
would be compelled to undergo in response to increase or decrease of 
the hemal areas. Furthermore, it would be still more difficult to explain 
why the injection masses become distributed irregularly in all directions 
among the cells of the parenchyma as soon as they pass out of the lacu- 
nae, and why the vein draining a node can never be shown to open directly 
into the subcapular blood space at the periphery, and can only rarely be 
shown to communicate indirectly with it through the intervention of 
blood spaces or blood islands in the interior. Such a supposition would 
require that the peripheral and central blood spaces must be thought of 
as extensive dilations which are practically coextensive in many cases with 
the node itself, and which communicate with the lacunae by very nar- 
row channels. It would then also be necessary for all the lymphocytes 
to pass through this wall of endothelium before they could enter the cir- 
culation ; and, on the other hand, the blood corpuscles would likewise 
have to penetrate the same wall upon entrance to and exit from the par- 
enchyma of the node; and the only alternative to this assumption is that 
both arteries and veins end freely in the parenchyma — for which I can 
find no evidence. I fully realize that the circulatory conditions as as- 
sumed above supply no reason for the migration of lymphocytes into the 
blood, and that they do not account for the occurrence of gaps in the 
walls of the lacunae. Yet the conclusions reached are no more invali- 
