1897] PLACE OF ISOLATION IN ORGANIC EVOLUTION 241 



able to preserving them ; for while cross fertilisation (amphimixis) 

 may stimulate variation, it also prevents the variations from 

 progressing by their mutual interference, and thus it tends to 

 keep a species constant, but ready to vary when circumstances 

 require it to do so. 1 Self- fertilisation, on the other hand, may be 

 unfavourable to the production of variations, but when one does 

 appear it has a good chance of being established. 



The general belief that breeding in-and-in is injurious has led 

 to the conclusion that a large and healthy progeny cannot arise 

 from a few parents if they are kept quite apart from all others. 

 But that cross-fertilisation is not necessary for the rapid increase 

 and continued health of the descendants from a few common an- 

 cestors is proved by the successful naturalisation of many animals 

 in New Zealand from very limited stocks. The honey-bee was 

 introduced by the Reverend Mr Cotton, chaplain to Bishop Selwyn, 

 who procured a few hives from Sydney; and, in 1S6G, wild bees 

 were common in the forests of the North. Island. Seven Chinese 

 pheasants were introduced in 1851, and six more in 1856 — more 

 than half being cocks ; and pheasant shooting near Auckland com- 

 menced in 1865. A few black swans were introduced by Sir 

 Walter Buller in 1864. A few cirl buntings were turned out 

 near Dunedin about 1868. A very few silver-gray rabbits were 

 released at Ivaikoura ; and, I believe, only three Tasmanian opos- 

 sums were turned out in the forests of Southland ; yet all these 

 species are now abundant and healthy. The herd of deer in the 

 Wairarapa (Wellington) has sprung from one stag and two hinds 

 turned out in 1863 ; and the herds in other parts of New Zealand 

 have all started from very few progenitors. Also many of the self- 

 introduced insects — as the English lady-bird {Coccinclla undecim- 

 j)v ' aetata), the drone-fly (Eristalis tcnax), the horse-bot {Gastropliilus 

 equi), and Lucilia cacsar, could only have been introduced in small 

 numbers, for each has spread from a single centre ; but yet they 

 have been very successful. It is true that several failures to 

 naturalise animals could also be given, but these failures do not 

 invalidate the evidence supplied by successful naturalisation, and it 

 is evident that, when the surroundings are favourable, it is quite 

 possible for a few individuals to give rise to a new and vigorous 

 species which might in time become dominant. 



Isolation must therefore be a true cause of the preservation of 

 variations, and also it must be an important one. The artificial 

 selection of animals and plants by man might just as well be called 

 artificial isolation. The breeder, or the horticulturist, certainly 

 selects the variation he wishes to preserve, but he also isolates it ; 

 and it is the isolation which causes the variation to be preserved — 



1 See Professor H. R. Orr's " Development and Heredity," p. 234. 



