140 BOTANY. [chap. iv. 



that the petal forming the labellum, intended as the 

 landing-stage for insects, appeared at the upper or 

 posterior part of the flower, and consequently proved 

 useless for the purpose intended, and although the 

 mistake was discovered, it appears that the line of 

 departure first taken could not be checked, and orchids 

 continue to grow the posterior petal as a landing-stage 

 and then, by a half-turn of the long, inferior ovary, 

 reverse the original position of the flower and bring the 

 labellum or landing-stage to the anterior or under side 

 of the flower, where it is of use to insects. In the great 

 majority of plants the pollen is dry and powdery, the 

 individual pollen-grains remaining distinct ; this is the 

 case in a few orchids, but in the great majority the pollen- 

 grains are stuck together in masses, two of which are 

 present in an anther which is situated above the stigma 

 in such a position that self-fertilization is impossible. 

 When a bee visits an orchid flower for the purpose of 

 obtaining the nectar situated in the spur that is a back- 

 ward continuation of the labellum, its head comes in 

 contact with the anther, and the two pollen masses 

 become attached to its head by a slender viscid stalk. 

 At the moment of removal the two pollen-masses stand 

 erect on the front part of the head of the insect, but in 

 a very short time bend forwards .and downwards, in fact 

 bend downwards until they are as much below their 

 original level as the stigma of the orchid flower is below 

 the stamen. When the ovules are ready for fertilization 

 the stigma becomes viscid, and a bee carrying the 

 pollen-masses on its head, on entering a flower in .this 

 condition, brings the pollen into contact with the viscid 

 stigma, to which it adheres. The mutualism between 



