448 Jlemarl's and Observations suggested 



while all the other flowers on the plant outside were deficient. 

 Plants that are full of full double flowers at one time, when the 

 plant is vigorous, will change and come more single when 

 checked by bad weather, or when the plant begins to ripen and 

 get woody. In the cultivation of dahlias also, as checking at 

 one time causes luxuriance afterwards to have more effect, the 

 roots intended to be flowered next year should be grown on 

 poor land. The buds will also be more perfectly formed in 

 well-ripened roots ; the crown of the root will always be more 

 perfect in a small well-ripened, than in a large soft root. It is a 

 mistake to suppose that the buds left on the cutting have any 

 effect on the root; these buds are always elongated into the stem, 

 and the root must form new buds for itself, which it will not do^ 

 unless ripened. To return to the raising of seedling double 

 flowers. The roses, pinks, carnations, and ranunculus change 

 the stamens only into petals, and sometimes these are only 

 partially so in very full flowers, and seed is comparatively easy 

 to be obtained from them ; we should, as before observed, select 

 from the fullest and best flowers. In the anemone the pistils 

 are changed into petals, the stamens unchanged ; seed of these can 

 therefore only be obtained from flowers not perfectly full, or by 

 impregnating flowers nearly single, with a tendency only to 

 fulness, with the anthers of full flowers. In stocks and wall- 

 flowers, both stamens and pistil are changed into petals : seed 

 cannot, therefore, be had from full flowers in these sorts; 

 and the only resource we have, is to save seed from those in 

 which a tendency to fulness has commenced, by having a petal 

 or two more than usual. In growing stocks from seed, they will 

 be moi'e likely to be double, if the plants are checked first by a 

 deficiency of nourishment, whether of water or manure, and 

 afterwards excited to luxuriance by a plentiful supply; and the 

 greater the change, the greater the likelihood of success. Old 

 seed, or seed dried, as in melons, gives a check ; we have had 

 instances of old neglected seed, which had been reckoned very 

 inferior when the seeds were fresh and new, come almost every 

 plant double, v.'hen a little had been left over and sown when 

 old. The seed for raising double flowers of any sort can hardly 

 be too old, if it will grow at all ; and the weak plants, first 

 stunted, and then luxuriated, will be found most successful: the 

 seed should be sown on heat, and the weak plants most cared foj*. 

 After flowers have once been produced double or full, the habit 

 of coming double will be retained, if kept so by rich cultivation. 

 When any variety has begun to sport, the plants should be raised 

 off those individuals which have not yet sported, as the sport- 

 ing habit might become fixed ; and this should be carefully 

 guarded against, by propagating from those roots that show the 

 fullest flowers. The double China asters, matricaria, feverfew, 



