ECOLOGICAL 191 



In spite of many experiments there remains much difference of 

 opinion as to the r61e of the various quahties of flowers in attracting 

 insect visitors. There seems to be, as should be expected, consider- 

 able diversity among the different types of insects, some being more 

 susceptible to colour and others to fragrance. To begin with, there 

 is often some experimenting on the insect's part, as has been noticed 

 in regard to newly introduced garden-flowers not previously tried. 

 The insects try and try again, and they sometimes enregister their 

 experience; as has been proved when humble-bees have discovered 

 how to bite a short-cut into the nectary of a new and difficult 

 bloom, and at once do so when they reach another flower of the 

 same kind. 



In the course of many tentative visits, some successful and others 

 disappointing, the insects estabhsh associations (to express it 

 psychologically) or conditioned reflexes (to express it physiologi- 

 cally) ; and these may be based on odours, brilliance, colour, and form, 

 to put them in their probable order of importance. Many insects live 

 in a smeU-world, even more than dogs do, and may have thousands 

 of olfactory bristles, disposed at strategic points on their body, on 

 the antennae in particular. When a successful worker-bee executes 

 her "honey dance" on the comb, the bystanders rush forward to 

 nose her, thus discovering a clue to the kind of blossom that it would 

 be profitable for them to search for. They do not, of course, reflect 

 over their behaviour, but it has been proved by Frisch that in a few 

 minutes they may discover the very patch of flowers where the first 

 bee — still in the hive — filled her honey-sac. When a worker-bee 

 finds a highly profitable flower she sometimes emits a little odorifer- 

 ous spray from a posterior gland, and this bee-odour on the plant 

 may be an olfactory cue useful to other bees, especially if the treasure- 

 trove be in itself scentless. The secretion of the odoriferous gland is 

 sometimes used as a signal in the confusion of swarming, or when 

 there is some difficulty in the return to the hive. There also seems 

 to be a characteristic queen-scent, the absence of which may be 

 rapidly detected by the workers of a hive which has lost its queen. 

 There is no doubt that scent counts for much in the life of bees and 

 many other insects; and there is thus good reason for attaching 

 importance to the experimentally verified conclusion that the 

 specific odour of flowers serves as a basis for association-forming. 

 But this must not lead us to forget the prior physiological question 

 as to the primary significance of odours in the plant's biochemical 

 routine. 



It has been proved by experiments, beginning with Lubbock's, 

 that bees and some other insects can distinguish colours as colours, 

 that is to say, apart from degrees of brilliance. When they have 

 established an association between a particular colour and a satis- 

 factory meal, they will pick out that particular colour from among 



