CORRESPONDING SOCIETIES 453 



things into their burrows. Darwin supplied them with triangles of paper 

 which could be pulled in easily by the tip, but with difficulty by the broad 

 base. He found that most of the triangles were pulled in the easy way, and 

 concluding that that involved a decision or a discovery that there was an 

 easy way, he was prepared to allow a modicum of intelligence to the 

 earthworm. 



Or let me show you, from an experience of my own, how simply an ex- 

 periment may develop. Twenty-four years ago I was spending a Saturday 

 afternoon basking in my garden in Edinburgh in October sunshine. I had 

 three papers, which I read at intervals. One was the Scotsman ; and as I 

 read I noticed that a small cloud of insects, the Winter midge {Trichocera 

 hiemalis), a relative of the Crane-flies, kept hovering above the newspaper- 

 sheet at a height of about 4 ft. Ultimately I laid aside the Scotsman, and 

 began to read the old green Saturday Westminster Gazette. The insect 

 cloud still hovered above, but I was surprised to see that it had descended 

 very markedly. I then tested the insects with Country Life, a magazine 

 with highly glazed paper, and the cloud at once rose, and hovered at a greater 

 height than even for the Scotsman. 



Very little consideration indicated that the height of the cloud of wmter 

 midges was regulated in some way by the colour or the light reflected from 

 the surfaces over which they hovered. The next step in the test was a 

 simple one. I prepared several pieces of cardboard all of equal size, and 

 painted them with different colours, black, white, blue, green, orange, red, 

 keeping the tone of the colours as nearly equal as possible. And then I 

 tried them on the winter midges ; but to make the decision as easy as pos- 

 sible for them I gave them only two colours to choose between at a time : 

 black and white, and the cloud hovered over the white ; I covered the black 

 card with a blue card, and the cloud still remained over the white : then I 

 covered the white with the red, so that they had to choose between blue 

 and red, and the cloud left the red and came within a minute to hover over 

 the blue. And so on for all the colours, until I had worked out the pre- 

 ferences of that hovering cloud : and I came to the conclusion from these 

 simple tests that colours influenced them in the order of their wave-length — 

 the shorter wave-lengths, blue end of the spectrum, were preferred to the 

 longer wave-lengths, red end of the spectrum ; indeed, that the insects were 

 simply reacting, not to any colour as a colour, but to intensity of reflected 

 light. 



Some of you are familiar with the interesting results that came of Lord 

 Avebury's equally simple experiments with bees and wasps and coloured 

 discs of paper. There is no end to that sort of experiment. I listened a 

 few months ago in Edinburgh to Miss Use describe her recent colour-tests 

 with butterflies. The results were extraordinary. Whenever it hatches, a 

 butterfly is attracted to a particular colour, but different kinds of butterflies 

 prefer distinctly different colours. Yet the original preference does not 

 hold a compelling attraction for a butterfly throughout its life, for when it 

 learns by experience the flowers which contain the nectar it desires, the 

 original preference fades before the colours of the nectar-bearing flowers. 



There are great possibilities in such simple ways of testing the reactions 

 of living things, and the method is one which lies ready to the hand of the 

 inquiring naturalist. Only I must warn you that experiments have to be 

 planned with care, and their results scrutinised with caution before a wise 

 conclusion may be ventured. 



Even so I do not pretend that those lines of observation ^yhich I have 

 suggested to you will lead to great discoveries, great discoveries lie in the 



