30O BOTANY [CHAP. X 



Letter 628 wonderful cases. The pollen of the closed imperfect cleisto- 

 gamic flowers differ in the transparency of the integument, 

 and I think in size. The latter point I could ascertain from 

 my notes. The pollen or female organs must differ in almost 

 every individual in some manner ; otherwise the pollen of 

 varieties and even distinct individuals of same varieties would 

 not be so prepotent over the individual plant's own pollen. 

 Here follows a case of individual differences in function of 

 pollen or ovules or both. Some few individuals of Reseda 

 odorata and R. lutea cannot be fertilised, or only very rarely, 

 by pollen of the same plant, but can by pollen of any other 

 individual. I chanced to have two plants of R. odorata in this 

 state ; so I crossed them and raised five seedlings, all of which 

 were self sterile and all perfectly fertile with pollen of any 

 other individual mignonette. So I made a self sterile race ! 

 I do not know whether these are the kinds of facts which 

 you require. 



Think whether you can help me to seed or better seedlings 

 (not cuttings) of any Melastomad. 



Letter 629 To F. Mliller. 



Down, March 2Oth, 1881. 



I have received the seeds and your most interesting letter 

 of Feb. 7th. The seeds shall be sown, and I shall like to 

 see the plants sleeping ; but I doubt whether I shall make 

 any more detailed observations on this subject, as, now that I 

 feel very old, I require the stimulus of some novelty to make 

 me work. This stimulus you have amply given me in your 

 remarkable view of the meaning of the two-coloured stamens 

 in many flowers. I was so much struck with this fact with 

 Lythrum^ that I began experimenting on some Melastomacese, 

 which have two sets of extremely differently coloured anthers. 

 After reading your letter I turned to my notes (made 20 

 years ago ! ) to see whether they would support or contradict 

 your suggestion. I cannot tell yet, but I have come across 

 one very remarkable result, that seedlings from the crimson 

 anthers were not -|Jths of the size of seedlings from the yellow 

 anthers of the same flowers. Fewer good seeds were produced 

 by the crimson pollen. I concluded that the shorter 1 stamens 



1 " Shorter stamens " seems to be a slip of the pen for " longer," 

 unless the observations were made on some genus in which the structure 

 is unusual. 



