THE WEAK SIDES OF NATURAL SELECTION. 81 



ccesiata) alike in the western highlands, and also in the S.W. of 

 Iceland. It is worth while to cumpare the two aeries in question. 

 The marbled appearance of the Scotch specimens so closely re- 

 sembling 1 their limestone or schistose rocks, and the dingy or 

 grimy appearance of the Icelandic ditto enabling them to lie perdu 

 on their native lava. While fully conceding that these instances 

 of similarity in insects (of which a hundred more examples might 

 be given) to the vegetable and the mineral world are ordered by 

 Providence as a safeguard against total or partial destruction by 

 their natural foes, I think we must be content to suspend our 

 judgment as to the particular agency by which this wondrous 

 similarity is effected. 



" Genesis of Species," p. 53. I am not personally quite certain 

 whether I thoroughly understand this term. By exposure of larva 

 or pupa to greater heat or more cold, or by feeding the larva on a 

 different food plant from that which it frequents in a state of nature, 

 we may obtain moths of different colour and markings, and by 

 breeding again from these and repeating the same experiments 

 through several successive seasons, we may perpetuate these super- 

 ficial distinctions, but can we so permanently perpetuate them 

 during the time we keep and register our observations of each suc- 

 cessive brood or during our own lifetime as to render it certain that 

 the insects, if restored to liberty and to their original food plants, 

 would not shortly or at any rate by degrees hai'k back to their former 

 tj r pe. The ultimate test of two true species is inability to pair with 

 one another, or at least of reproduction in a third generation, just as 

 the ultimate test of two genera is diversity of structure. Difference 

 of colour, size, markings, may frequently be noticed in the case of 

 two really different species, but these are not invariable nor final 

 tests, either of two different species, or two different genera of 

 butterflies, and even the two sexes of the same insect are often far 

 from presenting the same striking difference to those of another 

 tribe. 



