GEOGRAPHICAL DISTRIBUTION. 33 



stay some time longer in the enjoyment of their existing quarters. When 

 we consider, however, the return movement, at the end of winter, it is 

 doubtful, I think, whether scarcity of food can be assigned as its sole or 

 sufficient cause. But here we feel the want of knowledge. At present, we 

 are far too little acquainted with the physical peculiarities of those more 

 equatorial regions, which in winter are crowded with emigrants from the 

 north, to come to any final decision. It seems not too violent an assump- 

 tion to suppose that, though such regions are well fitted for the winter resort 

 of the bird population of the north, they may be deficient in certain neces- 

 saries for the nursery ; and it seems still less of an assumption to suppose 

 that even if such necessaries are not wanting, yet that the regions in ques- 

 tion would not supply food sufficient for both parents and offspring. But 

 another point must not be overlooked. The most sedentary of birds year 

 after year occupy the same quarters in the breeding season. It seems to 

 me, therefore, that among the causes of migration the desire of returning to 

 old haunts must be included." 



With regard to the distribution of reptiles and fishes, and their color- 

 variation, from our present standpoint of knowledge, we cannot as yet 

 apply the same generalization that governs the other classes ; but we do 

 know that the brightest colored individuals, and the largest generally of 

 these two classes, are found in tropical or semi-tropical regions. 



Unfortunately, reptiles particularly are rather restricted in their ranges, 

 and cannot move from point to point with the regularity and rapidity ot 

 birds, so that what we might apply to the class in general will not apply to 

 species, with but one exception as far as known, and for the reason that we 

 find distinct species inhabiting distinct localities, northernforms being replaced 

 by southern ones, with no intergradation of colors or varieties. The one 

 exception in the case of serpents is that of Opliibolus triangulatum, which, as 

 it proceeds southward, becomes darker and larger. 



With regard to the geographical distribution of insects, and their color- 

 variation thereupon depending, as yet the subject has not received the share 

 of attention that has been accorded to other classes, consequently we are 

 not in position to give any generalization regarding them ; but we may say 



as we have said of reptiles, that the largest and brightest colored species 

 3 z 



