208 THE STRUCTURE OF FLOWERS. 



long-styled form the stamens are unchanged, while the pistil 

 has elongated ; whereas, in the short-styled form, with now 

 elevated stamens, these and their pollen have presumably 

 become differentiated, while the pistil has remained un- 

 changed. Now the above result appears to indicate the fact 

 that the long-styled ])istil has not become physiologically 

 differentiated to so great an extent as the pollen of the long- 

 stamened form. The result is that it can be fertilised by the 

 unchanged pollen of the same form more easily than the short- 

 styled primitive form of pistil by the more highly differentiated 

 pollen. This is not stated as a proved fact, and must be 

 only regarded as a hypothetical suggestion. The extreme 

 limits of differentiation are reached when the flower is 

 heterostyled in form but dioecious in function. Thus 

 ^giphila ohdurata seemed to Mr. Darwin to be in a dioecious 

 condition, but derived from heterostylism, in which the 

 long-styled Avas apparently female, and the short-styled 

 male. 



The species which shows the most marked difference 

 between the produce of the legitimate fertilisation of the 

 two forms is P. Auricula (or cultivated vars. of Auricula). 

 It had been asserted by Prof. Treviranus that the long-styled 

 unions were absolutely barren.* Mr. Scott shows that this 

 idea arose from the fact that the plant in question had not 

 been crossed. His experiments prove that the short-styled 

 is the most fertile, whether legitimately or illegitimately 

 crossed, though in the latter the difference is slijjhter : in 

 the former the ratio being 8 to 6 ; and in the latter, 7 to 6. 



Homostyled forms of P. Auricula are not uncommon. Mr. 

 Scott found that 9 capsules gave 272 seeds, or an average 

 of 30 seeds per capsule. Compai^ing this with the following 

 results, its extreme fertility becomes apparent : — 



* Scott, I.e., p. 90. 



