208 SUPPLEMENT TO THE INTRODUCTION 
cation), I was at first more than surprised to hear of these experiments, for they seemed 
entirely to overthrow a view which I had up to that time considered an established 
oecological fact. But on careful consideration of these experiments, I came to the 
conclusion that Plateau’s inferences are not justified, and that another explanation is 
permissible. Let us take the experiment made with Digitalis purpurea. Plateau cut 
away not only the corolla-tube but also the style and stamens, till only a stump 
1cm. long remained (see Fig. 80). Gaston Bonnier (‘Les Nectaires, 1879, p. 61) 
observed many years ago that ‘les abeilles continuent 4 visiter en méme nombre 
les Digitales sur les pieds od toutes les couronnes avaient été enlevées” Plateau’s 
experiments confirmed this observation, for the visitors of uninjured flowers (Bombus 
terrester Z. and Anthidium manicatum Z.) also sucked the mutilated ones, finding it 
hard work to hold on while doing so—the resting-place presented by the complete 
corolla being absent. ‘Ainsi,’ says Plateau, ‘les hyménoptéres visitent encore, et 
d'une fagon effective, les fleurs de Digitales 
n’ayant plus ni leur couleur attractive, ni 
des dimensions les rendant trés visibles, ni 
la forme que ces animaux ont coutume 
d'utiliser pour parvenir aisément au nectar.’ 
But if we remember that a mutilated 
flower is an open cup, containing nectar 
which is constantly renewed from the 
base of the flower—where the nectar is 
situated—we shall realize that this nectar 
is freely displayed after removal of the 
corolla, and therefore—being exposed to 
the direct action of sunshine and wind— 
must evaporate more quickly, give out a 
stronger odour, and attract more strongly 
than when it is hidden at the bottom of 
a long corolla-tube. The visits of insects 
to this exposed nectar-cup would therefore 
FIG. 80. Digitalis purpurea, L. (after be more numerous than those to the 
Plateau). a, complete flower; 4, mutilated 25;2 
flower. uninjured flower, assuming that the corolla- 
tube has no significance as a means of 
attraction, No such observation, however, is found in Plateau’s account, so that the 
uselessness of the brightly coloured corolla for the purpose of attraction is not 
proven. 
Plateau has also made similar experiments to those on Digitalis with Lobelia 
Erinus, Oenothera biennis, Ipomoea purpurea, Delphinium Ajacis, and Antirrhinum 
majus. As before he removed the conspicuous part as far as possible, and yet the 
inconspicuous remnants—except in the case of Antirrhinum majus—received visits 
from insects sometimes almost as frequently as the complete flowers. Plateau 
explains that this is due to the fact that odour alone is the means of attraction. 
In my opinion the mutilated flowers should here again have received more numerous 
visits than the uninjured flowers, as in the former the odour of nectar must have 
been stronger. Since the mutilated flowers were not visited so frequently as those 
