788 REPORT — 1893. 



afford the slightest protection to an animal from its foes, a modification of colour, 

 which shall be assimilated to that of the surrounding country, is absolutely- 

 necessary. Hence, without exception, the upper plumage of every bird — whethea* 

 lark, chat, sylviad or sand-grouse —and also the fur of all the small mammals, and 

 the skin of all the snakes and lizards, is of the uniform isahelline or sand-colour. It 

 is very possible that some further purpose may be served by the prevailing colours, 

 hut this appears of itself a sufficient explanation. There are individual varieties 

 of depth of hue among all creatures. In the struggle for life which we know to 

 be going on among all species, a very slight change for the better, such as im- 

 proved means of escape from its natural enemies (which would be the effect of an 

 alteration from a conspicuous colour to one resembling the hue of the surrounding 

 objects), would give the variety that possessed it a decided advantage over the 

 tvpical or other forms of the species. ... To apply the theory to the case of the 

 Sahara. If the Algerian Desert were colonised by a few pairs of crested larks — 

 putting aside the ascertained fact of the tendency of an arid, hot climate to bleach 

 all dark colours — we know tliat the probability is that one or two pairs would be 

 likely to be of a darker complexion than tlip, others. These, and such of their offspring 

 as most resembled them, would become more liable to capture by their natural 

 enemies, hawks and carnivorous beasts. The lighter-coloured ones would enjoy 

 more or less immunity from such attacks. Let this state of things continue for a 

 few hundred years, and the dark-coloured individuals would be exterminated, the 

 light-coloured remain and inherit the land. This process, aided by the above- 

 mentioned tendency of the climate to bleach the coloration still more, would in a 

 few centuries produce the Galerita abyssinica as the typical form ; and it must 

 be noted that between it and the European G. cristata there is no distinction bwt 

 that of colour. 



' But when we turn to Galerita isabellina, G. arenic.ola, and G. macrorhyncha, 

 we have differences, notonly of colour, but of structure. These differences are most 

 marked in tlie form of the bill. Now, to take the two ibrmer first, G. arenicola has 

 a very long bill, G. isabellina a very short one ; the former resorts exclusively to 

 the deep, loose, sandy tracts, the latter haunts the hard and rocky districts. It is 

 manifest that a bird whose food has to be sought for in deep sand derives a great 

 advantage from any elongation, however slight, of its bill. The otlier, who feeds 

 among stones and rocks, requires strengtli rather than length. We know that 

 even in the type species the size of the bill varies in individuals — in the lark 

 as well as in the snipe. Now, in the desert, the shorter-billed varieties would 

 undergo comparative difficulty in finding food where it was not abundant, and con- 

 sequently would not be in such vigorous condition as their longer-billed relations. 

 In the breeding season, therefore, they would have fewer eggs and a weaker 

 progeny. Often, as we know, a weakly bird will abstain from matrimony alto- 

 gether. The natural result of these causes would be that in course of time the 

 longest-billed variety would steadily predominate over the shorter, and, in a few 

 centuries, they would be the sole existing race ; their shorter-billed fellows dying 

 out until that race was extinct. The converse will still hold good of the stout- 

 billed and weaker-billed varieties in a rocky district. 



' Here are only two causes enumerated which might serve to create, as it were, 

 a new species from an old one. Yet they are perfectly natural causes, and such 

 as I think must have occurred, and are possibly occurring still. We know so 

 very little of the causes which, in the majority of cases, make species rare or 

 common that there may be hundreds of others at work, some even moi'e powerfnl 

 than these, which go to perpetuate and eliminate certain forms " according to 

 natural means of selection."' 



It would appear that those species in continental areas are equally liable to 

 variation with those which are isolated in limited areas, yet that there are many 

 counteracting influences which operate to check this tendencj'. It is often assumed, 

 where we find closely allied species apparently interbreeding at the centre of their 

 area, that the blending of forms is caused by the two races commingling. Judging 

 from insular experience I should be inclined to believe that the theory of inter- 

 breeding is beginning at the wrong end, but rather that while the generalised forms 



