1862 i88i] MELASTOMADS 299 



yet. It adds immensely to interest of all such work as ours, Letter 626 

 and is indeed indispensable for much work. Experience, 

 however, has fully convinced me that the use of the com- 

 pound without the simple microscope is absolutely injurious 

 to progress of Nfatural] History (excepting, of course, with 

 Infusoria). I have, as yet, found no exception to the rule, 

 that when a man has told me he works with the compound 

 alone his work is valueless. 



To Asa Gray. Letter 627 



March 20th [1863]. 



I wrote to him [Dr. H. Criiger, of Trinidad] to ask him to 

 observe what the insects did in the flowers of Melastomacese : 

 he says not proper season yet, but that on one species a small 

 bee seemed busy about the horn-like appendages to the anthers. 

 It will be too good luck if my study of the flowers in the 

 greenhouse has led me to right interpretation of these 

 appendages. 



To J. D. Hooker. Letter 628 



Down, Nov. 28th [1871]. 



If you had come here on Sunday I should have asked you 

 whether you could give me seed or seedlings of any Melasto- 

 mad which would flower soon to experiment on ! I wrote 

 also to J. Scott to ask if he could give me seed. 



Several years ago I raised a lot of seedlings of a Melasto- 

 mad greenhouse bush (Monoch&tus or some such name) : 

 from stigmas fertilised separately by the two kinds of pollen, 

 and the seedlings differed remarkably in size, and whilst 

 young, in appearance ; and I never knew what to think of the 

 case (so you must not use it), and have always wished to try 

 again, but they are troublesome beasts to fertilise. 



On the other hand I could detect no difference in the 

 product from the two coloured anthers of Clarkia? If you 

 want to know further particulars of my experiments on Mono- 

 chatum (?) and Clarkia, I will hunt for my notes. You ask 

 about difference in pollen in the same species. All dimorphic 

 and trimorphic plants present such difference in function and 

 in size. Ly thrum and the trimorphic Oxalis are the most 



1 Monochcztum. 



2 Clarkia has eight stamens divided into two groups which differ in 

 the colour of the anthers. 



