1898) ANNULATE ANCESTRY OF THE VERTEBRATA 23 
monstrous, the end is gained by a waist (or, as in Scorpio, a perforated 
diaphragm, which is only a waist with the infolded external faces 
fused together). From the arrangement of the muscles, it appears 
that the narrow neck of this waist can be constricted when neces- 
sary. No such method of protection, of course, is conceivable in the 
case of the primitive vertebrates, in which a great part of the dorsal 
region of the body had to be protected from the ventral half of the 
same region. That protection was as necessary as it is in the case 
of the arachnids, we may surely believe if the rotundity of the body 
of the tadpole is any sort of repetition of the state of periodic dis- 
tension of our early ancestors, although of course in the tadpoles 
later modifications, such as the coiling of the intestine and the for- 
ward movement of the vent, are already superposed. 
The method of protection which actually was adopted—reading 
now from the embryological record—seems to have been as follows. 
A dorsal strip of the alimentary canal thickened and eventually 
separated off as the functions of the canal demanded free play to 
cope with the increasingly difficult digestive problems which the 
developing mouth and teeth, in their quest of new things to devour, 
continually sent down for solution. This thickening of a dorsal 
strip of the alimentary canal may perhaps again be referred to the 
downward pressure of the food at all times, whether the alimentary 
canal was distended or not. The dorsal epithelium would only be 
seriously pressed upon in the condition of actual distension, the 
ventral would be subjected to pressure whenever there was any solid 
food to rest upon it. 
It is conceivable that this dorsal strip of endoderm might have 
remained a protective plate if its sole function had been to screen 
the neuro-muscular system from the distension and churning activity 
of the alimentary canal. But, at the same time, it served another 
and almost equally important purpose which led to its stiffening 
longitudinally into an elastic rod. It is, indeed, horrible to con- 
template the possible fate of our unfortunate ancestors if, while the 
ventral half were fully distended with food, some sudden stimulus 
compelled the dorsal half violently to contract, as we know the 
modern leech can contract, to less than a quarter of its length. A 
stiff rod along the back alone could avert such a catastrophe. 
Hence I would suggest that the protective strip of endoderm 
stiffened longitudinally as it was progressively differentiated from the 
alimentary canal, and further, narrowed as it thickened, so as to 
permit of free serpentine movements of the body. We may then 
leave it as an elastic rod protecting on the one hand the spinal cord 
—which, as we shall see, must have been concurrently developed— 
from functional disturbance by the alimentary canal, and on the 
other, the alimentary canal from possible mechanical injury due to 
