SUMMARY AND CONCLUSIONS 155 



It seems difficult, in spite of carefully chosen expressions, to 

 make it clear that in my interpretation of mammalian lymphatic 

 ontogeny I am not trying to straddle the question of venous or 

 non-venous origin of the lymphatic vessels. What we ordinarily 

 describe as 'lymphatic channels' both in the embryo and in the 

 adult animal I am forced, by the results of my investigations, to 

 regard in amniotes as vessels developed absolutely independently 

 of the hsemal vascular system. The association of these channels, 

 in the mammalian embryo, with certain embryonal venous lines 

 is purely a secondary, mechanical and topographical relationship, 

 expressed by the condensed term of ' extraintimaP development 

 of mammalian systemic lymphatic vessels, and absolutely devoid 

 of genetic significance. This is, without reference to other ver- 

 tebrate classes, proved by the development within restricted areas 

 in the mammalian embryo of systemic lymphatic channels through 

 the direct confluence of intercellular mesenchymal clefts, not re- 

 lated topographically or in any other sense to the embryonal veins. 

 It is true that in the mammal this independent lymphatic genesis 

 is extremely limited, and that the majority of the lymphatic ves- 

 sels develop in close association with embryonal veins, as products 

 of the confluence of perivenous extraintimal spaces. But this is 

 merely, as shown by comparison with other amniote embryos, the 

 expression of the peculiar relations obtaining in the mammal be- 

 tween the venous and lymphatic circuits of the vascular system, 

 developed independently of each other. 



On the other hand, I am trying to give its full value to a hitherto 

 unrecognized component of the finished lymphatico-venous 

 organization of the mammal. I am trying to establish the ex- 

 tremely reduced and rudimentary mammalian lymphatico-venous 

 heart in the position which its phylogeny entitles it to occupy, as 

 the link or bond-piece between vertebrate lymphatic and venous 

 channels. I desire to make it as perfectly clear as it is possible to 

 do in the English language that this does not mean the acceptance 

 of a mixed, partly venous and partly independent genesis of the 

 lymphatic system as a whole. It merely calls for the morpho- 

 logical recognition of a distinct and valid structure, interDosed, as 

 the connecting link, between the definite venous channels and the 



