24 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



improve them, therefore, is to continue growing them from seed. 

 All do not ripen seed alike. A. puniceus, for instance, has its 

 fertile seeds blown all over the garden, and the seedlings, if not 

 weeded out, become a nuisance; but its contemporary, A. 

 s'pectabilis, I can never raise from seed, though the heads feather 

 .well. Aster seed is difficult to judge ; the achene, or body of the 

 seed, may appear very lean and shrivelled, and yet contain a live 

 germ, so proof must be waited for in the result of sowing. A. cor- 

 difolius, with its tiny flowers, ripens good seed in abundance ; so 

 do some of the forms of A. lavis and A. Novi-Belgii. Of others,, 

 the variety A bessarabicus (fig. 7) ripens seed better than its type^ 

 A. Amellus. Boltonia seeds freely, but, as far as I have tried, 

 reproduces the parent exactly. The middle-sized forms of 

 A. versicolor show a great disposition to vary and to improve. 

 A. Novce-Anglice has so many forms all well defined within the 

 species that it probably is capable of improvement by seed, and 

 might produce good accidental hybrids. Few gardeners seem 

 yet to have raised these plants systematically from seed with a 

 view to the improvement of the flower. I have often grown them 

 from bought seed, but the plants have either been worthless or 

 constant to type. It would be best to save a little seed from all 

 that ripens in gardens where many good lands are grown 

 together, and sow it each year. If a hundred gardeners out of 

 the ten thousand or more gardens where collections of Michaelmas 

 Daisies are cultivated would each devote a few square yards of 

 ground to raising every year a thousand seedlings, we should, 

 perhaps, get some advance annually. This would require hardly 

 any work ; the seedlings would be sown and left growing in the 

 open ground, and as nearly all flower the year they are sown, any 

 which showed merit would be marked and transplanted at once. 



There is no doubt that two adjacent Asters often form hybrids 

 in gardens, but good results are quite exceptional. Asters with 

 the best colour and largest flowers, A. spectabilis and A. pyr%- 

 iicbus for example, are over before the best seed-bearers begin, 

 and probably no one has tried saving pollen of these to fertilise 

 A. Novi-Belgii and A. cordifolius. An intermediate between 

 such different flowers as A. pyrenccus and A. cordifolius might 

 be excellent, but would more likely be worthless. Seedlings 

 come up plentifully amongst their parents in borders where 

 Asters are grown,- but I have found danger in leaving these 



