Raising Seedlings. 117 



line of demarcation ? There appears no limit to the field of labour. We have to learn 

 by actual experiment what can and what cannot be done. 



It should be known, in choosing varieties for this purpose, that the least double 

 kinds do not always perfect their seeds best. Such, upon less mature consideration, 

 might appear to be the case and has been asserted to be so, which error must have 

 arisen from the want of close observation. It does not depend so much on the 

 degree of fulness in a Rose as upon some other cause to me altogether inexplicable, 

 and not to be interpreted even by the acknowledged laws of the effects of hybridi- 

 zation, for some Hybrids seed freely whereas others are sterile, although of the 

 same origin and apparently similarly constituted. That the power of producing 

 perfect seeds does not depend on the degree of fulness may be established by the 

 fact that Pourpre Fafait, a mongrel-bred Bourbon Rose, and others, too full to 

 open their flowers at all times, ripen their seeds, although very many semi-double 

 varieties rarely form a seed-pod. That it does not depend on their being Hybrids 

 may be inferred from the fact that many of the Hybrid Chinese Roses, which are 

 decided Hybrids, seed freely. 



I have by the aid of the microscope examined numerous flowers with the view of 

 solving this difficulty. I have arrived at certain conclusions which may be considered 

 sufficient for practical purposes, or, what will prove still more important, may induce 

 others interested in the matter to push on the inquiry. 



The flowers were divided into three classes. The first class examined was that 

 which showed no disposition to seed, where the seed-vessels did not increase in size 

 after the falling of the petals. In this case I found the pistils huddled together, if I 

 may so express myself, and apparently sterile ; or, if not so, petals usurped the place 

 of the pistils and stamens, extending into the ovaria or seed-vessels. It was evident, 

 then, that such could not seed. 



In examining the next class, where there was a disposition to seed, where the seed- 

 vessels increased in size after the falling of the flowers but withered before arriving at 

 maturity, I found the pistils placed separately, and they appeared perfect and healthy, 

 but the stamens were either so few or so encased within the petals that the pollen 

 could not escape, and thus the flowers remained unfertilised. In some cases the 

 flowers were pendent, owing to which position and the relative length of the pistils and 

 stamens, the latter rising above the former, the pollen fell away from rather than upon 

 the pistils. In other cases, where the flowers stood erect, the pistils often rose above 

 the stamens, when the same consequences were likely to ensue. Flowers of this kind 

 will occasionally produce a pod of perfect seeds, which may be attributed to accidental 

 fertilisation, the conveying of the pollen by the insect tribe, or other causes. 



The next class taken in hand was that which ripened its seeds freely. 



The flowers here were found to have both stamens and pistils perfectly developed, 

 the former abounding in pollen, which, in a more advanced stage of the flower, was 



