184 The American Naturalist. [February, 
while the stomodeum joins the vesicle at its opposite end. The enteron 
is thus interposed between the anterior and posterior intestines, be- 
comes a part of the digestive canal and forms the mid-gut, which 
Crustacea almost completely lack.” The endoderm, hence, arises in- 
ternally from the protendoderm, and not by invagination. There is no 
gastrula among Arthropods, the apparent gastrula invaginations being 
really stomodeal invaginations. The alimentary tract arises from three 
rudiments in an entirely different way from the process followed in ani- 
mals with true gastrulation. Every little depression on the Arthropod 
blastoderm, M. Roule says, has been thought to bea gastrula when once 
the necessity for finding one was thought established by the early 
workers on the germ layers. However, gastrulation is not so important 
as the results of it, and M. Roule suggests that a very important differ- 
ence between Arthropods and other Ccelomates is the fact that their 
digestive tract isnot formed from any of the so-called gastrulas. 
(It will be noted that Heymons, in a recent paper, has also given up 
the idea of endoderm formed from a gastrula invagination, deriving 
the whole digestive tract in insects from the stomodeum and procto- 
eum. Korotoneff, too, has adopted Heymons’ view to a great ex- 
tent.) 
As development proceeds, the yolk enclosed by the enteric vesicle is 
gradually absorbed, the brain and ventral cord become differentiated, 
and the yolk mass in the body cavity is rapidly reduced. This reduc- 
tion is effected by phagocytes floating in the plasma, which washes 
the edges of the yolk. They eat it away until it is confined to the 
dorsal portion of the thorax, where it forms the hump mentioned 
above. Besides these wandering cells, other cells of the mesoderm 
elongate become grouped together and form muscle bands cris- 
crossing through the body cavity. By the end of this period all trace 
of metamerism has vanished from the mesoderm. The appearance of 
metamerism, noticed at an early stage, was due to collections of meso- 
derm cells at the bases of the successive pairsjof appendages, before 
they had pushed out sufficiently to accommodate those cells destined to 
shove in and form the inner structures of the limbs. The disappear- 
ance of this apparent metamerism in later stages is due to the seg- 
mental collections of cells having moved into the limbs to their definitive 
position. “This fact shows undoubtedly how the metameric disposi- 
tion of the ventral mesoderm is bound up in the distribution of the 
appendages in regular pairs as cause to effect. The first is the result 
of the second, and has no other value whatsoever.” 
In speaking of the external segmentation of the body, M. Roule says: 
