140 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
that in embryonic sheep, in addition to a postcava, an azygos vein also 
occurred. Rathke seems to have become cognizant of this fact, for in 
his third paper on this subjept he abandoned his earlier views on the 
fate of the postcardinals, and, without stating reasons for the change, 
adopted views more nearly in accordance with the new observations 
made by Stark. 
According to Rathke’s (’38, pp. 3, 4, and pp. 10, 11) later view, the 
anterior thoracic portion of the left postcardinal is involved in the for- 
mation of the hemiazygos, and the right postcardinal in a similar way 
enters into the formation, not of the postcava, but of the embryonic 
azygos. This view agrees in the main with the results obtained by recent 
investigators (Hochstetter, ’93, and Zumstein, ’96, 97) on other mam- 
mals, and is abundantly confirmed by our own observations on swine. 
The facts thus far stated are only a partial exposition of Rathke’s later 
opinion. Only the anterior portions of the azygos and hemiazygos 
are formed, according to Rathke’s later view, from the postcardinals, 
the posterior parts being developed from a system of longitudinal 
anastomosing trunks between the successive intercostal veins. These 
anastomosing trunks receive the blood from the intercostals, pass- 
ing it forward towards the heart, and thus form a longitudinal ves- 
sel, which gradually replaces a part of the original postcardinal. The 
extent to which this replacement occurs may be indicated as follows. 
The part of the hemiazygos extending from near the heart to the sixth 
intercostal vein represents a persistent part of the postcardinal, and the 
remaining part from the sixth intercostal posteriorly to the last one is a 
new formation from the longitudinal anastomosing vessel ; the part of the 
embryonic azygos from its connection with the heart posteriorly to the 
eighth or tenth intercostal represents the right postcardinal, the remain- 
ing posterior portion having been derived from the longitudinal anas- 
tomosing vessel of that side. After the right intercostals establish 
transverse connections with the hemiazygos, the azygos disappears, thus 
leaving the hemiazygos as a return trunk for the blood from the right 
as well as from the left intercostals. 
That the posterior portions of the azygos and of the hemjazygos in 
many mammals are new formations added to the remnants of the post- 
cardinals is now, we believe, generally admitted, and, as we ourselves 
have seen, is certainly true for swine ; but that these new formations, 
the accessory veins, develop from anastomosing branches between 
the intercostal veins, as stated by Rathke, is not, we believe, in accord- 
ance with fact. Of the several embryos examined by us at the stage in 
