898 MH. ^\. ^\'OODLAXD ox the [Nov. 27, 



with a highly developed renal cardinal mesh work: The encroach- 

 ment of the kidney tubules on the venous lumen necessarily 

 causes an obstruction to the flow of blood, notwithstanding that 



"the whole cardinal vein becomes greatly enlarged joari 



passu with the breaking up of its lumen into sinuses " (Shore, 6) in 

 its endeavour to maintain a suflicient passage* ; and it is the 

 failure of the enlarged posterior cardinal, in vertebrates above 

 fishes, to entirely effect this object, owing to the greater bulk and 

 complexity of the kidneys in these animals, which is, I believe, 

 the main cause of the adoption by the venous blood of an alter- 

 native course of flow. In fishes which possess no renal cardinal 

 meshwork, as e. g. some Selachians, and in those in which it is 

 not developed to a very great extent, as in most Teleosts and 

 Selachians, the whole of the venous blood is still able to return 

 to the heart via the mid-dorsal channel. But in Ceratodus 

 (apparently not in Proto2)terus, in which howevei' the hind- 

 limbs are vestigial), which approaches the Amphibia in general 

 structure, in Amphibia, Reptilia, and most Aves, the renal 

 cardinal meshwork becomes very much more developed, owing to 

 the greater size and complexity of the kidney in these higher 

 forms, and hence, the mid-dorsal channel becoming largely ob- 

 structed, the venous blood, particularly that returning from the 

 limbs in Ceratodus, Amphibia, and Reptilia, has to seek another 

 course for its passage to the heart. That other factors co-operate 

 to produce this mid-ventral channel is possible, and in some cases 

 probable, but the formation of the renal cardinal meshwork is I 

 think the chief. 



The principal other factor concerned in the production and 

 persistence of the ventral venous channel is one that has already 

 been referred to in my previous paper. I have mentioned that 

 the probable reason for the mid-dorsal route is the fact that in 

 the majority of cases it is in a more direct line with the large 

 caudal and iliac veins. Now, if the tail be unimportant and the 

 form of the body-cavity such as to render the mid-ventral course 

 as much in line with the posterior veins as the mid-dorsal, then, 

 even in the absence of a, renal cardinal meshwork, it is possible 

 for the blood to flow through a pelvic or coccygeo-mesenteric vein 

 as well as, or instead of, through the posterior cardinals. The 

 abnormal distribution of veins in Chelonia and Crocodilia has 

 already been thus interpreted. In the former group, it is probable 

 that the original presence of a fully-developed renal meshwork 

 aided in the production of a sub-abdominal course for the blood, 

 and that subsequently this course, being well adapted to the 

 circumstances, led in turn to the dissolution of the meshwork. 



* This is evident from the fact that " the cardinal sinus surrounding the nephri- 

 dial tubules maintains a large size as far back as tlie most posterior nephridium, 

 where the vessel almost suddenlj- narrows and passes into the thi'ee veins shown." 

 And again : " At this stage [of the Frog's development] the anterior parts of the 

 posterior cardinal veins are very large and dilated, and almost surround, as a vascular 



network, the pronephros The posterior parts, however, are more slender " 



(Shore. 6). 



