PEEFACE TO THE EEPEINT OF 1884. 



The text of the second edition has been left un- 

 touched, and I have merely given an account (which 

 does not pretend to be complete) of the progress of the 

 subject since 1880. 



Heterostyled Plants. 



C. E. Bessey (' American Naturalist/ June, 1880, 

 p. 417) made careful measurements of the corolla, sta- 

 mens, and style in a number of flowers of Lithospermum 

 longiflorum. He shows that the length of corolla, and 

 especially the length of the style, is very variable. An 

 appearance of dimorphism is thus produced; but meas- 

 urements of the pollen show that there is no real hetero- 

 stylism. 



C. B. Clarke (' Journ. Linn. Soc./ xvii. p. 159) has 

 made the curious observation that in Adenosacme longi- 

 folia the difference between the long- and short-styled 

 forms amounts to what would usually be called a char- 

 acter of first-rate systematic importance. In the short- 

 styled flowers, the stamens are on the corolla; in the 

 long-styled, they are at its very base, almost free from 

 it. In this form the corolla separates and leaves the 

 stamens standing on the ovary. 



