274 Practical Plant Biology. 



their succeeding generations, recording any changes which occurred 

 or noting their stability or immutability, we might then obtain 

 direct proof of transformation or support of the immutability of 

 the species observed. Such isolation experiments have been made 

 by many observers, but the results have not been decisive. In the 

 first case the time that has been possible to allot to them is prob- 

 ably vanishingly small compared to the time occupied in the 

 transformations of species in nature : then again, the changes which 

 have been noted have been either not transmitted or where they 

 have been inherited there has been the doubt that they are not 

 of sufficient magnitude to justify our regarding the new forms as 

 distinct species. 



Fortunately this experiment which is rendered unsatisfactory by 

 the time limit of human observation and indecisive owing to the 

 nature of the changes actually observed, has been carried out many 

 times over in nature. The most striking records of it are found 

 in the oceanic islands. These are islands separated from the 

 continents by large tracts of sea. They have originated, in more 

 or less remote periods of the past, from the outbursts of submarine 

 volcanoes, or from the upheaval of parts of the ocean floor. The 

 present population of plants and animals of these islands always re- 

 sembles the inhabitants of those continents from which they are most 

 accessible more closely than those of more distant lands. Hence it 

 is known that it is derived from chance colonists, borne on the air, 

 it may be as seeds or spores, on the sea, or even by birds. In cases 

 where isolation is very complete the periods of time elapsing be- 

 tween such colonisations may be very great. Hence the species 

 inhabiting these islands, derived from the continent from which 

 the islands are most accessible, develop there but slightly influenced 

 by crossing with the parent stocks from which they have been de- 

 rived. If species were immutable, the species now found on these 

 islands should be, so far as climatic conditions permitted, the same 

 as those on the colonising continent. On the other hand, if species 

 are mutable we would expect to find a larger or smaller number of 

 species on these islands not found on the continent, these new 

 species having arisen by transformation of some of the colonising 

 species. The proportion of new species would in this case depend 

 on the completeness of the isolation, the rate of transformation and 

 the period of isolation. We may illustrate the general relationship 

 found in these cases by a few facts drawn from Darwin's account 

 of the flora or plant population of the Galapagos Islands. 



This group of islands lies on the equator due west of Ecuador, 

 about 500-600 miles from the coast of South America. They are 



