IN SUMATRA. 229 



they have a singular knee extended into a fork-like projection, f, 



which in the flower lies just below the bright yellow antlieivs 



of the short stamens. The lower jjortion of the Vnv^ stamens 



takes a backward curve from the 



fork carrying the pores of its 



anthers far from those of the 



short stamens. This arranire- 



ment is most beautifully adapted, 



as was first pointed out by Fritz 



Miiller, for the cross-fertilisation FLo\vEn(piAGRAMMATic)or.-»ji:tA8TOMA. 



of the t)Lint ^^ si::cTiON.— p, petal; a\ aktmku 



-t ' * ^ ^ OFSHOiiT &TAMENS; /, FoKK OF I OM; 



Ihe bees invariably made stamens; n', antheu of lum; t>TA- 



lor the bnght yellow platform ./, stigma of fi^til ; 8, ovary, 

 offered by the bunch of short 



stamens (perhaps because they do not pt^rceive from a distance 

 the pink pistil and long stamens projected against the 

 pink corolla), and invariably received* \\\e pistil between their 

 legs, their feet settling also on the adjoining fork of the long 

 stamens. The instant effect of this is to collect the whole 



o 



depress 



stigma of the pistil 



o 



bee, while the 

 50 to the pores 



ot the long-stamened anthers) remains in constant contact 

 with its ventral side. At the moment of the bee's depar- 

 ture, the hooks on its feet, by pulling on the fork of the long 

 stamens, raise their anthers, bringing — now that there is no 



fear of producing self-fertilisation of the plant— their tips in 

 a collected bunch into contact with its sides an<l abdomen- 

 Long after I had made these observations, while working in the 

 laboratory of the Buitenzorg Gardens, Dr. Buntk pointed out to 

 me a fact of considerable imj)ortance which I was able to verify 

 for myself, that there was in very closely allied species of this 

 family a great difference in the shape of the pollen of the two 

 forms of anther; that uhile pollen of both shapes was fouiid on 



seemed fertile. We 



poll 



* O J. 



the short-stamened anthers. 



poll 



The reason why some organ of a 2>lant or animal has assumed, 

 as it were, an abnormal form, is not always easy to discover; 



