HARDWICKK S SCIENCE-GOSSIP. 



123 



time to select them. But, suppose we take this race 

 of bees, which thus accept the teaching of experience, 

 and set them down in a region where there are as yet 

 no blue flowers: what will happen? They will 

 learn to associate honey with white and yellow, or 

 whatever the colours may be. The result of this 

 will be — well, certainly, not to cause them to select 

 those flowers which began to show a shade of blue. 

 No, this bee will not do, and so Grant Allen brings 

 out his other bee. 



Suppose now a bee which has acquired, or been 

 gifted with a taste for blue, by any means, whatever, 

 except by an experience of blue flowers. Set it down 

 among a race of flowers of another colour, which 

 are beginning to acquire a blue tinge. Suppose it 

 perseveringly selects the blue, and you get an 

 explanation of the origin of blue flowers, which 

 satisfies some people. This is the bee by means of 

 which Grant Allen evolves his Monkshood ; the 

 orgin of its taste remaining a mystery. When it is a 

 question of explaining the origin of blue flowers you 

 use this bee ; when you wish to account for the bee's 

 taste for blue you use the other. It is an important 

 point that when the one is in use the other should be 

 kept out of sight ; and a similar method is very 

 largely used when the principles of natural selection 

 are brought down from the general to the particular ; 

 from the plausible assertion of principles to the 

 difficult task of explaining the origin of any 

 particular organism. 



In my original paper I brought forward the case of 

 the periwinkle, and the common scarlet poppy as 

 markedly blue and red flowers not frequented by 

 bees. Along with other facts, I considered it to 

 show that bees do not prefer blue and red to other 

 colours. This was met by the statement from 

 Miiller that Vinca minor is visited by seven species of 

 bees, and Papaver rhesus by seven also. In reply, 

 I pointed out that the number of individual bee- 

 visits, and not the nnmber of species, was the 

 important point. This again is met by the supposition 

 that in these cases the number of visits is roughly 

 proportioned to the number of species. And this is 

 apparently supposed to take away the force from my 

 argument. I can show, however, that Miiller's 

 observations on the above supposition lead to the 

 same conclusion that I drew from my own. 



With regard to Vinca minor there is evidently 

 something requiring explanation, for while Miiller 

 states that he has seen it abimdantly visited by 

 insects ; he also informs us that other celebrated 

 observers — Sprengel, Darwin, Delphino, and Hilde- 

 brand — seem never to have seen any insects upon it 

 but thrips. On Vinca major, which differs from 

 Vinca minor chiefly in size, Miiller once saw 

 Bomhis agroru?n. But, accepting the seven species 

 seen on Vinca minor as representative, the record, 

 when compared with the same observer's notes of 

 visits to white, yellow, and green flowers, is not good 



enough to show that bees prefer blue. For example, 

 of white flowers, the inconspicuous chickweed, 

 Stellaria media, has six' species, and Lamhnn album 

 sixteen ; of yellow flowers, dandelion has fifty-eight, 

 and Ramcnculns ficaria eight ; of green flowers, the 

 raspberry has eleven, and the gooseberry nine 

 species. The blue Linum usitatissimum has a record 

 of two. 



As to the poppy, Midler's list shows it to be much 

 more frequently visited than I had supposed ; but 

 when we compare it with the first of a nearly-related 

 yellow flower, Chelodonium majus, we find the 

 number of species the same, viz., seven. Therefore 

 the case of Papaver rhceas supports the view that 

 bees do not prefer red. It would be interesting 

 if other observers would give their results as to 

 these flowers. 



But while I have endeavoured to show that, taking 

 the number of species as roughly proportional to the 

 number of individual visits,, Miiller's observations 

 point also to the conclusion that bees do not prefer 

 blue and red, I am far from thinking the method 

 may not, in many cases, be very misleading, as I 

 have before hinted. Many of Muller's species of 

 bees are doubtless rare, 'and it can hardly be supposed 

 that his lists are complete. This appears evident 

 when it is noted that he records only two species 

 as visiting the lime, Tilia Europcea. No one, surely, 

 would be disposed to deny that this flower receives 

 more bee-visits than chickweed or dandelion. If, 

 however, the number of individual visits could be 

 taken, I believe the same conclusion would be 

 pointed to. 



With regard to my observations — why does Mr. 

 Tansley, by-the-way, call them experiments ? — on the 

 comparative number of visits to Veronica Buxbatunii 

 and Stellaria media, I find that Muller's records are 

 quite similar ; taking, as before, number of species 

 as our criterion. Unfortunately he does not mention 

 V. Buxbaumii, but, perhaps, V. chajnadrys will do 

 as well. To compare with the five species of Stellaria 

 media, it has a record of four, whilst V. spicata 

 has two. 



As to Mr. Tansley's distrust of my observations 

 I shall say nothing ; it is quite fit and proper that 

 he should prefer Sir John Lubbock's experiments. 

 Doubtless, as he suggests, my results are capable of 

 being "explained away." 



And, by-the-way, it is Grant Allen who speaks of 

 Sir John Lubbock having taught the bees. " Sir 

 John put drops of honey on slips of glass above 

 coloured paper ; and when he had once taught a bee 

 to feed from one slip, say the blue, he found that it 

 would return straight to that slip, even when the 

 relative places of the colours had been transposed."* 

 But I also hope it was not really a matter of 

 " teaching." 



* "Knowledge," April 14th, 1883, p. 508. 



G 2 



