COMPARATIV.E ANATOMY. 161 



bills ; the inside of tliem, towards the edge, being thickly 

 set 'with parallel or concentric rows of short, strong, sharp- 

 pointed prickles. These, though they should be called teeth, 

 are not for the purpose of mastication, like the teeth of quad 

 rupeds ; nor yet, as in fish, for the seizing and retaining ol 

 their prey ; but for a quite diiierent use. They form a filter. 

 The duck by means of them discusses the mud ; examining 

 with great accuracy the puddle, the brake, every mixture 

 wliich is likely to contaui her food. The operation is thus 

 carried on : the liquid or semiliquid substances in which 

 the animal has plunged her bill, she draws, by the action of 

 her lungs, through the narrow interstices which lie between 

 these teeth, catching, as the stream passes across her beak, 

 whatever it may happen to bring along with it that proves 

 agreeable to her choice, and easily dismissing all the rest. 

 Now, suppose the purpose to have been, out of a mass of 

 confused and heterogeneous substances, to separate for the 

 use of the animal, or rather to enable the animal to separate 

 for its own, those few particles which suited its taste and 

 digestion ; what more artificial or more commodious instru- 

 ment of selection could have been given to it, than this nat- 

 ural filter ? It has beer, observed also — what must enable 

 the bird to choose and distinguish with greater acuteness, as 

 well probably as what greatly increases its luxury — that the 

 bills of this species are furnished with large nerves, that 

 they are covered with a skin, and that the nerves run down 

 to the very extremity. In the curlew, woodcock, and snipe, 

 there are three pairs of nerves, equal almost to the optic 

 nerve in thickness, which pass first along the roof of the 

 mouth, and then along the upper chap down to the point of 

 the bill, long as the bill is. 



But t(5 return to the train of oui observations. The 

 similitude between the bills of birds and the mouths of quad- 

 rupeds is exactly such as, for the sake of the argument, 

 might be wished for. It is near enough to show the contin- 

 lation of the same plan ; it is remote enough to exclude the 



