1 30 E VOL UTION OF TO-DA Y. 



where the digestive canal ought to be, and thus by 

 its simple bulk effectually prevents the regular for- 

 mation of these organs. The embryo does seem to 

 try, even here, to follow the primitive history, for it 

 develops a mouth in a roundabout way, and separ- 

 ates off cells which correspond to the digestive tract. 

 But the mouth frequently closes up, since it is of no 

 use, and the presence of the bulky food prevents 

 the alimentary canal from being formed. After the 

 food is used up the mouth again appears, the ali- 

 mentary canal becomes functional, and the typical 

 history is once more resumed. In other words, the 

 embryo attempts to follow the line of ancestral his- 

 tory, but the presence of food makes this impossi- 

 ble, and as a result some of the history is skipped. 

 If now it be possible to find another annelid in 

 which no food is stored up in the egg, evidently 

 these modifying conditions will be absent, and the 

 embryology can follow more closely the ancestral 

 history. In this way two annelids closely related 

 might differ radically in their embryology, without 

 this fact at all affecting the truth of the hypothesis 

 that embryology tends to repeat past history. 



Or take a still more striking illustration. Accord- 

 ing to modern ideas derived from anatomical and 

 paleontological evidence, the ancestors of the birds 

 were at one time aquatic animals, breathing by 

 means of gills somewhat as do fishes. If, there- 

 fore, embryology repeated exactly the past history, 

 we should expect to find gills in the embryo chick. 

 But consider the condition in which the chick passes 

 its embryonic life: closed within the egg-shell, un- 



