224 PROFESSOR FRANK J. COLE 
blood corpuscles occur in uninjected specimens in all the so-called lymphatic spaces of 
the body. It has been stated that the vessels rupture and fill the lymphatics in bring- 
ing the animal up from the bottom. This may conceivably be so at Cullercoats, where 
Myzxine are not usually found within the 23-fathom line, and most of my material was 
collected in over 25 fathoms of water; but the Californian Hag used by Jackson occurs 
quite commonly in from 10 to 20 fathoms of water. If, therefore, the presence of 
injection mass in the lymphatics is due to rupture of the vessels, we are driven to 
distrust any results produced by that method. It will, however, be obvious from the 
preceding discussion that in this matter JACKSON is mistaken. 
We have seen that in Myxzne blood enters the pleural or peribranchial sacs via 
the vascular papille on the afferent and efferent arterial system, although chiefly by 
the former. If this be the case, there must be some corresponding connection with 
the venous system by which the contents of the sacs may be drawn off. Some chance 
injections enabled me to demonstrate this connection in a very striking manner. 
Injection by the ventricle of the heart is a little difficult. ‘The heart is tucked 
away under the anterior lobe of the liver, and in getting at it some vessels are certain 
to be cut, and the injection bleeds. I therefore determined to try an injection by the 
portal vein, which is very easy to get at and to ligature, and to expose which only a 
very slight operation is necessary. ‘lo my great surprise, the result was a complete 
injection of the veins of the whole body, with apparently nothing whatever in any of 
the arteries. The liver and gall bladder are injected first (and presumably also the 
anterior portal vein). ‘Then the sinus venosus fills up from the liver, and finally the 
injection travels back along all the veins, which become well filled. Examination of 
the specimen afterwards with the dissecting microscope revealed a little injection in 
the auricle, still less in the ventricle, a trace in the cardiac aorta, and none in the 
afferent branchial arteries or in the gills. Nevertheless the pleural sacs were full of 
injection, and it had also passed freely into the large sub-dermal sinus. In this case 
there was and could be no question of rupture, since the injection had first to pass 
through the liver before reaching the sinus venosus. Therefore the fact that the 
pleural sacs filled up in a case where the veins were also well injected, and the possi- 
bility of injection reaching the sacs wa the arteries is absolutely out of the question, 
shows that the sacs must at some place be.in communication with the venous as well as 
the arterial system. I have not specially investigated this connection, but I believe it 
to be with the superior jugular veins in the neighbourhood of the heart. 
A brief note by Kirinckowstrém * may be quoted in confirmation of the above :— 
“Auch im Innern des Korpers bestehen weit ausgedehnte Lymphriiume, die sich iiber — 
und unter dem Cisophagus, zwischen den iiusseren und inneren Zungenmuskeln iiber 
dem Gaumen und rings um das Nasenrohr erstrecken. Wie oben erwihnt, stehen sie 
in weitoffener Verbindung mit den subcutanen Lymphsicken. Auch die Kiemensicke 
sind von zahllosen grésseren und kleineren Lymphriumen, die alle mit einander in 
* Biol, Foren. Férhand., Bd. iv., 1891. 
