SECT. XIII THE TRILOBITES 227 



first antenticTS being simply sensory cirri ; and secondly 

 because in no other group of early Crustaceans do 

 the anterior antennae show any traces of ventral para- 

 podia as masticatory ridges. In Eurypterus, where 

 the method of life we attribute to the animal would 

 certainly have developed them into jaws had it been 

 possible, they almost entirely disappear. On the other 

 hand, we have examples of the second antennae 

 developing their ventral parapodia as masticatory 

 ridges, not only in Limulus but also in Eurypterus. 



We are thus disposed to complete Walcott's resto- 

 ration by adding a small pair of anterior antennae 

 on each side of the labrum, and a pair of posterior 

 antenna, developing, probably as their most important 

 part, a pair of jaws strong enough, if not to crush and 

 destroy, at least to push food into the opening of the 

 oesophagus ; whether the sensory part was developed 

 or not is not so easy to decide. 



It may be noticed that it was not so necessary for the 

 Trilobites to have large crushing jaws under the 

 labrum, as the masticatory ridges of the first trunk 

 limbs were, as in the Eurypteridae, highly developed 

 to function as chief mandibles. The crushed food 

 would have to be forwarded towards the opening 

 of the oesophagus, and then pushed in by special 

 appendages at the sides of the opening. This 

 point is almost as interesting from a biological as 

 from a morphological point of view. We have 

 already had two entirely different combinations of 

 head appendages as jaws. In Apus, the third and 

 fourth head limbs form the mandibles (or chief 



q 2 



