MR. Gt. L1EWI8 OTf JAPANESE LAWaURIID^. 353 



ofdrouglit. Witt the drought the temperature rises, the sun 

 and wind scorch the foliage and upper parts of the tree or 

 plant, while a porous soil below drains the roots. The thick part 

 of the tree or palm still retains a certain moisture, and, stimu- 

 lated by the heat, continues to grow. An environment of this 

 kind would cause a proneness to enlargement above the roots 

 in many vegetables existing within its area ; and as long as 

 these conditions were fulfilled, this phenomenon would be a 

 noticeable feature in the landscape. What I have roughly deli- 

 neated here is a process which is, I believe, the origin of the 

 bulbous forms of vegetation, whether it be that of a hyacinth, an 

 onion, or a stunted oak. But who would doubt that, if the climate 

 of this area changed, became, as it were, more like Ceylon, with 

 continuous rain, this bulbiform habit which I have depicted would 

 disappear, and in time be eradicated altogether by the plants 

 resuming a more constant and uniform activity in all their parts. 

 The dryness of this area at one period of the year is the illus- 

 tration of the " creative " and " preservative " cause I wish to 

 notice. The periodical dryness after luxuriant rain which ori- 

 ginated the bulbous habit must continue to recur, otherwise it 

 disappears. A small Scilla on Wimbledon Common is subject to 

 these dry conditions as I describe in August ; and in Japan, the 

 home of the Liliaceae, there is a long dry winter season (Ent. 

 Mon. Mag. xviii. pp. 5-7) , which has a great effect there on both 

 animal and vegetable life . 



The above illustration may serve to show what " reversion," 

 commonly so called, is to my mind. All ideas of reversion ap- 

 pear to me to be greatly at variance with the doctrine of evo- 

 lution. But let us suppose that an actual Ceylonese climate 

 takes possession of the area spoken of What happens ? The 

 plants would resume their growth, as I have suggested, in all 

 their parts, and their globosity would disappear. Yet we could 

 not call this reversion ; for in reality there is no retrograde move- 

 ment, although the second condition of these plants may be more 

 like those forms from which they originally arose, or are sup- 

 posed to be derived. 



Sometimes entomologists, when they observe a succession of 

 likenesses, such as those here traced in plants, in animals 

 widely separated from each other in a system of classification, set 

 the phenomenon down as originating through " natural selection " 

 producing a resemblance which is supposed to be " protective," 



