332 
WILLIAM F. ALLEX. 
distal or growing areas of tlie larger mesenchymal cavities 
apparently furnishes us with the solution of the origin and 
nianner of growth of these mesenchymal spaces and cavities. 
Here in section may be seen the remains of what has tran- 
spired. The protoplasmic processes of the mesenchymal cells 
in the centre were found to be disintegrating and increasing 
in diameter on the border of the cavity^ while some of the 
enclosed mesenchymal cells had increased in size, and were 
becoming spherical to form the red corpuscles, and others in 
the walls of the larger cavities had flattened in places to form 
a part of the lining for the future caudal heart. 
Another possible source of a very small portion of the 
caudal hearts in Polistotrema is shown in the 20 mm. 
series A to be from the proximal ends of three embryonic 
intersegmental veins (Fig-. 7, HP.s.F.). These embryonic vessels 
have already been described under the head of intersegmental 
veins, and no trace of these vessels was found in the 20 mm. 
series B or any of the later series until the 85 mm. series was 
reached, indicating that the persisting intersegmental veins 
of the adult did not arise in the caudal heart region until a 
stage midway between (30 and 85 mm. The anterior and 
posterior of these so-called embryonic intersegmental veins in 
the 20 mm. series A (Fig. 7, Infs. V.) are in every way com- 
parable to the intersegmental veins that are forming anterior 
and posterior of the caudal heart region. Tracing them fi-oni 
the dorso-lateral surface of the left caudal vein they pass 
laterad for some little distance, and then bend dorsad to end 
blindly, about on the level with the left caudal artery. If, then, 
these are true embryonic intersegmental veins, arising as 
sprouts from the left caudal vein, it would be possible for 
their proximal portions to contribute slightly to the forma- 
tion of the left caudal heart. Concerning the middle vessel, 
which has been described as an embryonic intersegmental 
vein anastomosing with a corresponding intersegmental artery, 
is shown in transverse section (Fig. 14, hits. V.) to bend dorsad 
much sooner than the other embryonic intersegmental vein 
did. It is possible, then, that this vessel is an embryonic 
