II Pitcher Plants 25 



use for reference. Even more useful is it to excavate ancestral 

 tombs, and bring out their fossils ; indeed none has set 

 us a better example than Darwin himself in all these very 

 ways. 



But not the smallest living scene — not even this bee upon 

 its flower — is to be understood from our museums and her- 

 baria : for this exhaustive division of labour, with its ento- 

 mological and botanical specialists, in winning extension of 

 exact and detailed special knowledge, had lost sight of de- 

 veloping that vague general knowledge with which childhood 

 begins. Watching the bees among the flowers is an old and 

 happy occupation, not only for children but their elders, and 

 many a writer, from the prosaic economist up to the master 

 of all poets, has long ago said his say ; their points of 

 view, however, were alike always too impressionist or too 

 anthropomorphic in standpoint to contribute anything to- 

 wards exact natural science, and so mummy-labelling and 

 shelving went on indoors undisturbed. Once only a hundred 

 years ago a childlike old German botanist went out into 

 the garden and watched summer after summer, till he saw 

 what the bees were doing, all unconsciously, to the flowers, 

 and learned how they and the flowers were fitted one to the 

 other in every detail of form like hand and glove ; and 

 when he was sure of his facts he could keep the secret no 

 longer ; he noted down everything that he had seen, and 

 this too with excellent drawings, calling his book, in naive 

 childlike delight and pride, The Secret of Nature Dis- 

 covered / But the botanists indoors would not look at his 

 book, save at most to say — What nonsense ! what childish 

 fancies ! what waste of time ! So it was soon forgotten, 

 and lay for a century unnoticed, until a naturalist, not con- 

 ventionalised in the museums, nor over- educated for his 

 intellect, but persistently childlike in questioning and 

 watching, and watching and questioning again, should 



