XXV] DARWIN 5 



they reject. This may be obtained in two ways. Those who 

 keep insectivorous birds, such as thrushes, robins, or any of 

 the warblers (or any other that will eat caterpillars), may 

 offer them all the kinds they can obtain, and carefully note 

 (i) which they eat, (2) which they refuse to touch, and (3) 

 which they seize but reject. If the name of the caterpillar 

 cannot be ascertained, a short description of its more pro- 

 minent characters will do very well, such as whether it is 

 hairy or smooth, and what are its chief colours, especially 

 distinguishing such as are green or brown from such as are of 

 bright and conspicuous colours, as yellow, red, or black. The 

 food plant of the caterpillar should also be stated when known. 

 Those who do not keep birds, but have a garden much fre- 

 quented by birds, may put all the caterpillars they can find 

 in a soup plate or other vessel, which must be placed in a 

 larger vessel of water, so that the creatures cannot escape, 

 and then after a few hours note which have been taken and 

 which left. If the vessel could be placed where it might be 

 watched from a window, so that the kind of birds which took 

 them could also be noted, the experiment would be still more 

 complete. A third set of observations might be made on 

 young fowls, turkeys, guinea-fowls, pheasants, etc., in exactly 

 the same manner. 



" Now the purport of these observations is to ascertain the 

 law which had determined the coloration of caterpillars. 

 The analogy of many other insects leads us to believe that 

 all those which are green or brown, or of such speckled or 

 mottled tints as to resemble closely the leaf or bark of the 

 plant on which they feed, or the substance on which they 

 usually repose, are thus to some degree protected from the 

 attacks of birds and other enemies. We should expect, 

 therefore, that all which are thus protected would be greedily 

 eaten by birds whenever they can find them. But there are 

 other caterpillars which seem coloured on purpose to be con- 

 spicuous, and it is very important to know whether they have 

 another kind of protection, altogether independent of disguise, 

 such as a disagreeable odour and taste. If they are thus pro- 

 tected, so that the majority of birds will never eat them, we 



