368 THE BIOLOGY OF FLOWERING PLANTS 



the Ophrys flowers are avoided in this way on their own 

 spikes ; they are simply not visited. The result for O. 

 muscifera and O. aranifera is that very little seed is set. 

 Many counts have been made by Darwin, Detto, and 

 Eckhardt, and the number of capsules with seed ranges from 

 less than i to about 17 per cent., and usually lies between 

 5 and 10 per cent. ; O. apifera is completely fertile, but 

 this is due to the fact that it is self-pollinated. In view of 

 our new knowledge of the colour sense of the bee, it seems 

 very likely that the reason for the lack of visits is that the 

 dark red blossom is invisible or indistinguishable from 

 green. It seems that we are here dealing with three flowers 

 which have definitely disadvantageous characters ; the 

 resemblance to insects is quite accidental and visible only to 

 us. It is dangerous to use such a resemblance as the basis 

 of hypotheses as to supposed advantages to the plant. This 

 danger is illustrated even better by another example recently 

 described. Mobius (19 12) states that the flowers of certain 

 species of Delphinium, e.g. D. elatum, have a very strong 

 resemblance to a flower with a bee entering ; it is even 

 stronger than in Ophrys apifera. But bees visit the Del- 

 phinium industriously, and the bee-like appearance is 

 explained as being useful. " Perhaps the bees are tricked 

 to believe that others of their species have visited the flowers, 

 and that they are therefore worth visiting ; so that they 

 then search for unvisited flowers on the same stalk, and on 

 coming closer are convinced that the flowers are still free 

 and realise their error." The author's use of the word 

 " perhaps " is well justified. 



Scent. — The importance of scent is peculiarly difficult 

 to estimate. Experimental work is difficult, because there 

 is no objective method of classifying scents or of measuring 

 their strengths. We can distinguish between " aromatic," 

 " sweet," and " nauseous " odours in a rough sort of way 

 (Linnaeus set up seven classes on such a basis, and Henning 

 (1916, see also Parker, 1922) has proposed a similar 

 classification), and we can distinguish between strong and 

 weak by direct comparison, but that is all. It is certain 



