226 ANNALS OF THE CARNEGIE MUSEUM. 
and secondly by the fact that the structure of the marsupium in the 
Unionide becomes highly specialized, or, to express it concisely, 
this family makes a special effort to bring the marsupial apparatus 
to the highest degree of efficiency. All differentiation is connected 
with two purposes: the lengthening of the breeding season, and the 
change of the discharge of the glochidia from a ‘“‘natural”’ to an “un- 
natural’’ manner, if such an expression may be allowed. As regards 
the first, the subfamily of the Unionine is as yet in an undifferenti- 
ated condition, possessing a short breeding season (being tachytictic), 
without specialization.4 But in the Anodonting and Lampsiline the 
breeding season is extended over the winter and the glochidia, after 
they are fully developed, are not discharged immediately, but retained 
for a long period in the marsupium (bradytictic). This renders it 
necessary to develop special devices in the marsupium, and the most 
urgent need apparently is to provide the necessary oxygen for the 
glochidia enclosed in the marsupia. It is now interesting to observe 
how this purpose is accomplished in two different ways by the two 
subfamilies. In the Anodontine, the lateral, secondary water-tubes 
cut off from the central ovisac, undoubtedly have the purpose of 
keeping up a lively current of water around the swollen marsupial 
mass. Nothing similar to this is known in the Lampsiline, but in 
the case of these the whole marsupium bulges out beyond the original 
edge of the gill, and this bulging mass is enclosed in a rather thin 
membrane, favoring osmotic processes. Further, there is a tendency 
to locate the marsupium in the posterior part of the gill, and to push 
it toward the lower posterior end of the shell, so that it is close to the 
branchial opening, where fresh and pure water enters the animal. 
In addition, a number of the Lampsiline develop special papille and 
flaps on the edge of the mantle, just at the place toward which the 
marsupium is pushed, and these structures surely have the purpose 
of producing a lively current of water over the marsupium. Further- 
more, the thin membrane enveloping the protruding part of the 
marsupium, and its position near the branchial opening, are apparently 
connected with the peculiar discharge oi the glochidia in the Lamp- 
siline, which is through the edge of the marsupium, through holes 
4 Haas (1910e, p. 19) comparing the marsupium of Anodonta and Unio expresses 
the opinion that that of Anodonta is more primitive than that of Unio, which is 
entirely erroneous: just the opposite is the case, that.of Anodonta being much more 
complex and specialized. As long as views like this prevail, we cannot expect to 
arrive at a proper understanding of the system of the Najades. 
