176 
description, although accurate in many respects, is inadequate. I have 
therefore reworked the whole subject of the blood-vessels of the 
lymphatic gland. The aim has been to begin with the artery and to 
follow all its branches until they reached the histological units of the 
gland. Then the course of the capillaries has been followed through 
the follicles and the lymph cords to the veins and the latter were 
traced to the point where they leave the gland. It was found that 
every unit has its artery in the centre and its vein at its periphery 
much as are the relations of the artery and veins in the villus of the 
intestine and in the lobule of the lung. 
The two large mesenteric lymphatic glands situated one on either 
side of the anterior mesenteric artery of the dog were used in this 
investigation. Since they receive their blood supply from the anterior 
mesenteric artery which is easy to expose, by ligating its branches all 
of the pressure of injection may be exerted upon the blood-vessels of 
the glands. 
Simple injections of an aqueous solution of Prussian blue were 
made at first to show the general plan of the vessels. Afterwards 
injections of carmine gelatine were made from the artery filling com- 
pletely the capillaries of the follicles followed by ultramarine blue gela- 
tine which does not pass through the capillaries. The blood vessels 
were also in a number of instances injected with cinnabar gelatine 
from the artery and with an aqueous solution of Prussian blue from 
the portal vein. 
The glands were hardened in alcohol and sectioned. Carmine 
staining was used for the tissues injected with blue and cinnabar, and 
hematoxylin for the carmine injections. These staining methods distinctly 
outlined the follicles and cords without obscuring the injected vessels. 
In most instances serial sections were made of the gland in directions 
both longitudinal and transverse to the follicle. This enabled me 
to follow the vessels from section to section in the same follicle. In 
tracing the course of the larger vessels thick sections were used, while 
for the finer vessels thinner sections had to be employed. 
A number of branches from the anterior mesenteric artery enter 
the gland at its hilus, which pursue a more or less direct course 
through the hilus stroma toward the medullary substance (Fig. 1 A). 
Most of the arteries on leaving the stroma substance enter the trabe- 
culz in which they run some distance before entering the medullary 
substance. An artery on leaving a trabecula to enter the medullary 
substance "may send a small artery (Fig. 1 E) through the trabecula 
to the periphery of the gland to anastomose with the arteries in the 
