324 BOTANY [CHAP. X 



Letter 647 accurate. It is a golden rule, which I try to follow, to put 

 every fact which is opposed to one's preconceived opinion in 

 the strongest light Absolute accuracy is the hardest merit 

 to attain, and the highest merit. Any deviation is ruin. 

 Sincere thanks for all your laborious trials on Passiflora. I 

 am very busy, and have got two of my sons ill I very much 

 fear with scarlet fever ; if so, no more work for me for some 

 days or weeks. I feel greatly interested about your Primula 

 cases. I think it much better to count seed than to weigh. 

 I wish I had never weighed ; counting is more accurate, 

 though so troublesome. 



Letter 648 To J. Scott. 



Down, 25th [1863 ?] 



From what you say I looked again at Bot. Zeitung} 

 Treviranus speaks of P. longiflora as short-styled, but this 

 is evidently a slip of the pen, for further on, I see, he says 

 the stigma always projects beyond anthers. Your experi- 

 ments on coloured primroses will be most valuable if proved 

 true. 2 I will advise to best of my power when I see MS. 

 If evidence is not good I \vould recommend you, for your 

 reputation's sake, to try them again. It is not likely that 

 you will be anticipated, and it is a great thing to fully 

 establish what in future time will be considered an important 

 discovery (or rediscovery, for no one has noticed Gartner's 

 facts). I will procure coloured primroses for next spring, 

 but you may rely I will not publish before you. Do not 

 work too hard to injure your health. I made some crosses 

 between primrose and cowslip, and I send the results, which 

 you may use if you like. But remember that I am not 

 quite certain that I well castrated the short-styled primrose ; 

 I believe any castration would be superfluous, as I find all 

 [these] plants sterile when insects are excluded. Be sure and 

 save seed of the crossed differently coloured primroses or 



1 " Ueber Dichogamie," Bot. Zeit., Jan. 1863. 



2 The reference seems to be to Scott's observation that the variety 

 rubra of the primrose was sterile when crossed with pollen from the 

 common primrose. Darwin's caution to Scott was in some measure 

 justified, for in his experiments on seedlings raised by self-fertilisation 

 of the Edinburgh plants, he failed to confirm Scott's result. See Forms 

 oj Flowers, Ed. II., p. 225. Scott's facts are in the Journal Linn. 

 VIII., p. 97 (read Feb. 4th, 1864). 



