v CHANGE FROM A REPTILE TO A BIRD 55 
heredity and variation. In form and character the off- 
spring take after their parents, but in almost every case 
there is some slight discernible difference. The in- 
dividuals that have variations that fit them better for 
life survive, those that have injurious, or, in some 
cases, those that have only useless variations or none 
at all, perish, Thus are produced new species, one 
useful variation after another being accumulated by 
Natural Selection. The sickly and those who are 
unsuited to their surroundings have no chance, for 
the law is mercifully ruthless. 
The facts that I have stated seem to prove much, 
The struggle for existence is indisputable, and evolu- 
tion through Natural Selection seems almost beyond 
dispute. But when we come to investigate more in 
detail how it has acted, we are met with great diffi- 
culties. A living man of science has said that for the 
explanation of the brilliant colours of the butterfly 
Darwin’s theory is but a barren formula. It may 
be that only his own imagination is barren. Later 
on, in a chapter on colour and song, I hope to 
show that colour is, at any rate, connected with 
Natural Selection. The particular difficulty that 
confronts us when we try to trace the evolution of 
birds is that the development of wings would have 
been useless, and worse than useless, unless accom- 
panied by other changes. And just as he who builds 
a Latin verse conceives a master stroke and puts into 
his line, that before was tame and commonplace, a 
purpureus pannus, then suddenly, to his dismay, finds 
that his brilliant emendation has ruined the grammar 
and the sense, so we may imagine a reptile, that 
