260 BOTANY [Chap. X 



Letter 590 i| hour, I found an indusium with hairs on the outer edge 

 perfectly clogged with pollen, and horns protruded, which 

 before the 1^ hour had not one grain of pollen outside the 

 indusium, and no trace of protruding horns. So you will see 

 how I wish to know whether the horns are the true stigmatic 

 surfaces. I would try the case experimentally by putting 

 pollen on the horns, but my greenhouse is so cold, and my 

 plant so small, and in such a little pot, that I suppose it 

 would not seed. . . . 



The little length of stigmatic horns at the moment when 

 pollen is forced out of the indusium, compared to what they 

 ultimately attain, makes me fancy that they are not then 

 mature or ready, and if so, as in Lobelia, each flower must be 

 fertilised by pollen from another and earlier flower. 



How curious that the indusium should first so cleverly 

 collect pollen and then afterwards push it out ! Yet how 

 closely analogous to Campanula brushing pollen out of the 

 anther and retaining it on hairs till the stigma is ready. I 

 am going to try whether Campanula sets seed without insect 

 agency. 



Letter 591 To J- D - Hooker. 



The following letters are given here rather than in chronological order, 

 as bearing on the Leschenaultia problem. The latter part of Letter 591 

 refers to the cleistogamic flowers of Viola. 



Down, May 1st [1862]. 



If you can screw out time, do look at the stigma of the 

 blue Leschenaultia biloba. I have just examined a large bud 

 with the indusium not yet closed, and it seems to me certain 

 that there is no stigma within. The case would be very 

 important for me, and I do not like to trust solely to myself. 

 I have been impregnating flowers, but it is rather difficult. . . . 



I have just looked again at Viola canina. The case is odder : 

 only 2 stamens which embrace the stigma have pollen ; the 3 

 other stamens have no anther-cells and no pollen. These 2 

 fertile anthers are of different shape from the 3 sterile others, 

 and the scale representing the lower lip is larger and differently 

 shaped from the 4 other scales representing 4 other petals. 



In V. odorata (single flower) all five stamens produce 

 pollen. But I daresay all this is known. 



