884 The American Naturalist. [October, 
connected axially with the inner layer, and a primitive streak in 
the front part of which the three layers.are connected axially, and 
in the hinder part of which the middle layer is connected with 
the outer layer only. This stage is quite well known; cf, 
Heape, zz, on the mole; Bonnet on the sheep, 6; Kölliker on 
the rabbit (Grundriss); Selenka on the opossum, 37 ; Lieberkühn, 
28, and others; especially the very careful descriptions of the 
rabbit's layers by C. Rabl, 29. 
Now, we do not yet understand the homologies of the mam- 
malian blastodermic vesicle, hence we cannot explain the pecu- 
liar relations of the mesoderm to Hensen’s knot, as the homology 
of the knot is unknown. However, since mammals are in all 
respects related to the Sauropsida, and especially since there is a 
close likeness between the subsequent stages of the two’ classes, 
it is probable that the origin of the mammalian mesoderm will 
be shown ultimately to be essentially the same as in reptiles. At 
present it seems to me impossible to offer any satisfactory 
interpretation of the observed double origin of the mammalian 
mesoderm. 
The Vertebrate Type of Origin of the Mesoderm—The preceding 
paragraphs show that in all classes of vertebrates the origin of the 
mesoderm is essentially the same, except in the mammals. The 
relations in the mammals we do not understand. In the non- 
mammalian vertebrates the mesoderm first appears as a thicken- 
ing of the entoderm over a ‘not inconsiderable area around the 
concrescing blastodermic rim, and it becomes separated from the 
entoderm by the gradual parting of the upper cells to form the 
true mesoderm from the lower cells or permanent entoderm ; this 
delamination does not take place next the blastodermic rim (or 
after concrescence in the axial line), hence in the region of the 
primitive streak the three layers may be connected for a time; 
further in the prolongation of the axis in front of the streak the 
mesoderm does not separate from the entoderm, thus forming the 
head-process. It is important to note that the mesoderm arises 
over a considerable area during the same period ; that its forma- 
tion may be more or less advanced before concrescence of the 
rim; and that after concrescence it stretches across the axis of 
