310 
WILLIAM F. ALLEX. 
between 15 and 20 nnn. From 20 to 30 mm. and from 48 to 85- 
mm. my series are good and fairly complete. 
My own specimens were all fixed in Tellyesnicky’s potas- 
sium-bicliromate-acetic mixture, embedded in paraffin, cut 
transversely 10 microns thick, and stained in either Heiden- 
hain^s iron or Hansen’s ha^matoxylin and counter-stained with 
a saturated alcoholic solution of orange G plus a little acid 
fnchsin. From my experience with the injection method I 
am satisfied that in many cases it would have only demon- 
strated the later stages of development, namely, the stage 
where the mesenchymal spaces had united to form a continuous 
canal, unless perchance there had been an extravasation of 
the injection mass, breaking down the mesenchyme separating 
these cavities, and possibly going still further. Upon inject- 
ing the lymphatics of fishes I have frequently broken down the 
delicate walls separating the lymphatics from a vein, filling 
the veins just as one sometimes fills the veins in the liver 
upon injecting the bile-vessels. In well-preserved material 
there is no reason whatever for regardino- the meseuchvmal 
spaces found in the region of growing lymphatics as artifacts, 
for if such, they would occur elsewhere in the mesenchyme. 
My previous studies on the lymphatics of fishes have indi- 
cated that this system is more closely allied to the venous 
system than is the case in the higher vertebrates. This is 
especially true of Polistotrema, where these vessels always 
contained blood, although it should be stated that no direct 
connections were found with the arterial system. The primi- 
tive vertebrate doubtless had no distinct lymphatics other 
than the veins, which would also function for lymphatics, a 
portion of which in their later phylogeny may have differ- 
entiated into lymphatics. If this be true, it is reasonable 
to suppose that the lymphatics have the same origin, onto-^ 
genetically, as their more primitive ancestoral veins, of which 
considerable data shows that the larger veins do not arise as 
sprouts from a central venous centre, but rather grow from 
periphery to the centre, very much after the manner of the 
formation of the blood islands, mesenchymal spaces, in the 
