334 



NATURE 



[Feb. lo, 1881 



the honey ; and they will fly from flower to flower, passing 

 over the less attractive kinds. 



The genus Rhinanthus has been made especially inte- 

 resting by Dr. Miiller. Rhinanthus is essentially a "bee- 

 flower," but R. alpinus has been modified so as to be 

 fertilised by Lepidoptera. The ordinary entrance by 

 which bees visit the flowers of Rhinanthus cristagalli is 

 here closed, and a special " butterfly-door," a minute 

 aperture at the tip of the upper lip, has been developed ; 

 it is moreover advertised to Lepidoptera by a pair of violet 

 flaps on each side of the entrance. The interesting point 

 about this genu; is the existence in it of a species which 

 shows in what manner the flower of R. nlpiuus (fitted for 

 Lepidoptera) may have been developed out of a"bee-flower'' 

 such as 7v'. cristagalli. This intermediate form [R. alccto- 

 rolophns) possesses a " Lepidoptera-door " like that of 

 R. alpinns, but has not closed the bee-door; it is therefore 

 visited by both bees and Lepidoptera and cross-fertilised 

 by both. In spite of our knowledge of this interesting 

 intermediate form, the evolution of R. alpinus remains a 

 difficulty. For although it is adapted for the legitimate 

 visits of Lepidoptera only, it is plundered by bees, who 

 break in by the closed bee-door ; and these useless 

 or injurious visits are actually more frequent than the 

 advantageous ones of Lepidoptera. It seems impossible 

 to believe that a butterfly-flower could be developed 

 under such circumstances, and the only explanation which 

 Dr. Miiller offers requires the assumption of two changes 

 of condition. First, the spread of the bee-fertilised 

 ancestors into regions (such as the Heuthal in the 

 Bernina) where they would be visited exclusively by 

 Lepidoptera, and where the present form of the corolla 

 might be developed. Secondly, we must suppose that 

 the plant has spread into regions where it is visited by 

 bees ; or else the plant has remained stationary while 

 bees have invaded its habitat. A similar kind of argu- 

 ment is applied to those flowers, Polygala alpestris, 

 various Papilionaceffi, &c., which, though structurally 

 adapted for fertilisation by bees, are, in the alps, chiefly 

 visited by Lepidoptera. They cannot therefore have 

 been developed in their present habitat, but must have 

 spread to the alps from the lowland regions.' Dr. Miiller 

 compares flowers like Rhinanthus alpinus, Gentiana, 

 (Cyclostigma), Erica carnea, &.C., to air-breathing verte- 

 brates which have been derived from water-breathing 

 ancestors, whose gills have been replaced by special air- 

 breathing structures — the lungs ; while Rhinanthus alecto- 

 rotophus corresponds to those intermediate forms which 

 still possess both gills and lungs. We may perhaps, by 

 an inversion of the simile, compare such plants as 

 Polygala alpestris, a nearly unmodified bee-flower visited 

 almost exclusively by Lepidoptera, to the Cetacea, which, 

 though actually breathing air, lead almost the life of a fish. 



Dr. MuUer's treatment of the genera Rhinanthus and 

 Gentiana are instances of the manner in which many 

 other groups are treated. Thus the interesting series of 

 forms exhibited by the Caryophylleas suggest the possible 

 steps through which bright-coloured flowers adapted for 

 Lepidoptera, such as those of the pinks, have been deve- 

 loped from the pale-coloured scentless flowers with 

 unprotected nectar which are chiefly visited by Diptera. 



' This view is, for reasons given in the text, put fonvard merely as a 

 speculation (p. 559). 



The present volume gives continual evidence of Dr. 

 Muller's knowledge of the structure and habits of insects. 

 But it does not (and this could not have been expected) 

 contain anything like the valuable study of insects con- 

 tained in the author's " Befruchtung " — a research of 

 which we take this opportunity of expressing our high 

 admiration, in which we shall be joined by those of our 

 readers who remember the excellent articles by Dr. Miiller 

 on insects which appeared in these pages. 



In considering the modifications of flowers produced by 

 their relations to insects, we are prepared to find that, for 

 instance, flowers fertilised by bees differ in shape from 

 those visited exclusively by Lepidoptera, but it does not 

 seem prima facie probable that the colours should be 

 characteristic of the two classes. Yet Dr. Miiller believes 

 that this is the case, and shows how it may probably be 

 connected with fundamental differences between the lives 

 of bees and butterflies. A bee having to work not only 

 for its own livelihood but also for its nest, is driven to a 

 greater degree of activity than the self-indulgent butterfly. 

 It is therefore important that a bee should work with 

 more method, and thus it happens that bees usually visit 

 one species of flower at a time, instead of passing from 

 species to species and wasting time in the constant change 

 of action. On this account it is obviously an advantage 

 for a bee to be able to distinguish easily a large number 

 of species, thus their unconscious selection has acted in 

 the direction of producing great variety of colouring. It 

 is indeed a remarkable fact that flowers which are visited 

 by short-trunked insects are often characterised by a 

 single colour (usually yellow or white) running through a 

 whole group, whereas closely-related " bee-flowers " are 

 generally varied in colour. 



Here then we have a curious chain of cause and effect, 

 beginning with the fact that bees have to provide food for 

 their young, and ending with the varied colours of species 

 of Labiates, Pedicularis, and Trifolium, &c. ! If any proof 

 is needed of the correctness of the first link in the argu- 

 ment, it may be found in the curious fact that parasitic or 

 cuckoo bees differ markedly from other bees in their 

 habits,^ visiting merely those flowers whence they can 

 obtain enough honey for themselves with least trouble 

 and doing it in a dawdling manner which meets with no 

 approval from Dr. Mijller. 



In the flowers adapted for or chiefly visited by Lepi- 

 doptera, red and in a less degree blue are the prevailing 

 tints. There seem to be some grounds for believing that 

 butterflies prefer flowers resembling themselves in tint. 

 Thus in sunshiny weather the orange-yellow flower-heads of 

 Arnica, Scnecio Doronicum, Sec, and the orange-red ones 

 of Crepis aurea and Hieracium aurantiacum are veritable 

 "Tummelpliitze" for yellow-red species of Argynnis and 

 Melitaa. On the other hand the blue Phyteumas which 

 decorate the alpine turf in thousands are especially visited 

 by blue Lepidoptera (" Blues "). It is hard to say whether 

 the butterflies have preferred flowers coloured like them- 

 selves, because these tints have been already rendered 

 attractive through sexual selection. Or whether vice versd 

 we may suppose that the colours of their favourite flowers 

 have reappeared as sexual decoration ; or lastly, that 



' P. S22 ; Dr. Miiller adds a caution that the number of observations on 

 this point are perhaps hardly sufficient to warrant a well-grounded con- 

 clusion. 



