PREFACE. XV 



in the two forms, but the extreme difference in the 

 point of insertion of the stamens." A mid-styled form 

 exists having a short pistil and short stamens seated on 

 the same level, only a little way up the tube of the 

 corolla. Mr. Clarke adds that heterostylism is quite 

 common in the Coffee tribe. Mr. Hiern, in his obser- 

 vations on the Eubiaceae of tropical Africa (' Journal 

 Linn. Soc. Bot./ vol. xvi., 1877, p. 252), remarks that 

 dimorphism occurs commonly, or at least in some spe- 

 cies, in four or five genera in the tribe of Hedyotideae. 

 Mr. M. S. Evans states (' Nature,' Sept. 19, 1878, p. 

 543) that in Natal there is a heterostyled Eubiaceous 

 plant, which occasionally, though rarely, presents a 

 third form, and in this the pistil and stamens are of 

 equal length and both are exserted from the mouth of 

 the corolla. He adds that he has found four other 

 heterostyled dimorphic plants, and one of these is a 

 monocotyledon. 



Lastly, I have given (p. 135) Bouvardia leiantha as 

 doubtfully heterostyled; Mr. Bailey has now sent 

 me dried specimens, and as far as the lengths of the 

 pistil and stamens are concerned the species is clearly 

 heterostyled; but no difference could be detected in 

 the size of the pollen-grains; so the case must remain 

 doubtful. 



With respect to trimorphic heterostyled plants, Dr. 

 Koehne, who has described the Lythraceae of Brazil, 

 has, with great kindness, sent me a long account of 

 them. He knows twenty-one species which are hetero- 

 styled, and 340 which are homostyled. He informs 

 me that Lythrum thymifolia is not heterostyled, and 

 that I must have received some other species under 

 this name. There are many dimorphic species in 

 America. Pemphis acidula is distinctly dimorphic, so 

 are some species of Eotala and Nesaea; thus two new 



