320 



NATUKAL SELECTION. 



[Cii. XXXVII 



more 

 at SI 



may 



perhaps be chiefly guided by 

 his ability to trace in the individuals inhabiting all the 

 intermediate hei 

 of the series to another. 



a gradual passage from one extreme 



Intercrossing of 



varieties 



eficial. — It would be 



xperiment 



made^ to cross individuals taken from the lowest station with 

 those hardier races which have been formed by acclimatisa- 

 tion in the upper regions of the mountain, and ascertain 

 whether they would produce as much seed as individuals 

 fertilised by the pollen of plants of the 



same 



If 



com 



it would afford an indication 



comm 



nature of that character which distinguishes wild species 

 from artificially formed races. There is good reason, how- 

 ever, to believe that before any difficulty of crossing, or any 

 deficiency of prolific power in the offspring, would be appa- 

 rent, the races must depart so widely from each other that 

 their distinctness as species would already be a debateable 

 question with the naturalist. And this brings us to the 

 principal obstacle which we encounter when we endeavour to 

 refer the gradual formation of a new species to variation 

 and natural selection. If some degree of sterility was found 

 in the offspring of slight varieties, and this want of prolific 

 power went on augmenting in proportion as the deviation from 

 a common stock became more and more marked, the fact that 

 closely allied species inhabiting the same region keep distinct 

 would be intelligible. But the phenomena are precisely the 

 reverse. Instead of any reluctance being exhibited by shght 

 variations to intermarry and propagate their kind, their in- 

 termixture, on the contrary, takes place freely and infuses 

 fresh vigour and fertility into the species. Individuals of 

 the norinal type are always the most numerous, and slight 

 varieties are usually soon merged in the general averao-e, so 

 that the new characters disappear. In some cases where 

 the races are so wide apart as to be thought by some 

 to belong to distinct species, it is only necessary to cross 

 their mongrel or hybrid offspring with pm^e individuals 



t 



/ 



C^' 



^ 





ti 



lias 



be 



above 



and cl 

 ieti 



only d 



ill eacl 

 and p( 

 seem t 

 sucli ai 

 still iu« 

 pelaga 

 must I 



before 

 and ho 

 spring 



Breei 



that 

 kind : 



CC] 



i 0, 



u. 



breedi 



Tliei 

 eenturie 



Tankerv 

 number 



fertile tb 



Vrica, 

 '^casioii. 



T e de,; 



t- 



b: 



is 



^01. 



II 



