310 CORAL AND ATOLLS 



would have been suitable for their earlier stages. The failures 

 from this cause must be beyond estimate. 



The case of the dragonflies gives one aspect of the picture 

 of Nature's methods, her absolute prodigality, her countless 

 failures, and her implicit reliance upon chance. But there is 

 another aspect, and this is the absolute success of her methods. 

 My other case illustrates this. 



Some while before I left the atoll we introduced tomatoes 

 into our garden, and the tomatoes came as seeds in packets 

 from England. With the first crop of plants a swarm of the 

 larvse of a moth, new to the atoll fauna, began to devour the 

 tomato leaves. No examples of this insect had been seen 

 before, and yet the suitable food plant, within a very short 

 time of its appearing above the ground, had received its proper 

 tenants. There can be no question that the insect was 

 certainly not introduced as eggs among the seeds, for, apart 

 from the general improbability of such an occurrence, it is a 

 species which is not found in England, although it is widely 

 distributed over the world. With peas the same story was 

 repeated, and the moth made abundant by their introduction 

 is again a very widely distributed one. 



The history of the tomato and the pea is probably one of 

 universal application, for it is almost true to say that of all 

 the vegetable waifs which have arrived in the islands, ever}' one 

 has been appropriated by the insect whose particular choice it 

 happens to be. 



We have seen, in the case of the " Queensland bean," how 

 uncertain are Nature's ways of vegetable colonising, and we 

 have seen, in the case of the dragonflies, the enormous 

 chances against successful insect invasion ; — it then becomes a 

 matter of very real wonder that almost every chance vegetable 

 colonist has received, by chance, its appropriate animal client. 



The question of finding a suitable food plant does not 

 affect all insects equally, for some are far easier to accommodate 

 than others. The larvre of such moths as Prodenia littoralis 

 will eat almost any low-growing plant, and will make them- 

 selves at home practically wherever vegetation of any kind 



