AND CROSS-BREEDING. 165 



with what Danvin has observed in his paper on the existence 

 of two forms in the genus Linum, where, in summing up the 

 good gained by the inevitable crossing of the dimorphic flowers, 

 and numerous other analogous facts, he says that these all lead 

 to the conclusion that some " unknown law of nature is here 

 dimly indicated to us." And this law, when discovered, may 

 disclose more mysteries, tending, perhaps, to the wider diver- 

 gence of species, with constitutions and habits better fitted for 

 the climates and localities in which they may be cast, as well 

 as for subserving the purposes they are intended to fulfil in the 

 economy of nature. In looking at Rammculacecz, with their 

 innumerable male and female organs (and the same thing 

 occurs in the Myrtacea, most of the Rosacece, some of the 

 Hypericacea, and in many other families and tribes), the idea 

 was long ago suggested to me, that each separate row, from the 

 outer to the inner circle of the stamens, might have some sepa- 

 rate function, just as I believe that the long and short stamens 

 have their separate functions ; and with the view of testing the 

 matter, I had last summer begun experiments with these outer 

 and inner stamens; but, other aims and objects interfering, 

 I gave up the experiment after I had begun it on these 

 Clematises. 



But to make success certain, it is my custom, as I have 

 already stated, in crossing any of these polyandrous flowers, to 

 take the entire bloom of one kind, and lightly to brush over, 

 with all its anthers, the stigmas of the flower to be crossed, and 

 leave nature to make her own selection. In referring to the 

 Rubus tribe and its species, I am reminded of an intention I 

 expressed in my former paper of perhaps returning to them 

 afterwards. I again experimented upon them last summer. 

 But though I tried various crosses among them, and recipro- 

 cated the cross, I had no success in any, except between the 

 R. biflorus and the R. Idaus, and that only where I made the 

 latter the seed-bearer. And to make sure of either event 

 success or failure I had the R. Idczus early potted and put 

 under glass, emasculating every bloom I meant to cross ; and 

 for more security I stripped off all other flowers nay, more, 

 I put the emasculated flowers under fine gauze bags, to ward 

 off the invasion of insects. When ripe for crossing I removed 

 the bag, and, on effecting the cross, I replaced it. In this way 

 I succeeded in ripening three berries of the cross R. Idaus by 

 R. biflorus, of which I sowed the seed between the 5th and 

 1 6th July, though as yet none have vegetated. But R. biflorus 

 stubbornly rejected a reciprocal cross. Again I tried both of 

 these on R. rupestris, and the latter on them ; and though R. 



