Un the Culture of the Dahlia. 495 
I should say it will not. — Never having tried the experiment, I may 
be mistaken, and should be glad to be set right. I lately saw a 
grower who recommended me to sow all the dross and lightest seeds 
I could collect, as being the most likely to produce fine double 
flowers •. — this is quite reversing the question, yet I feel inclined to 
tall in with the opinion. As all double flowers are monstrosities, it 
would appear probable that degenerate, crippled, and drossy seed may 
produce them, whilst the stout, bold, ripe, and perfect, would all pro- 
duce much stronger plants, more in their natural habit, and conse- 
quently more likely to have single flowers. On this point I would 
observe, that I have almost invariably found this year, that if any 
plant among my seedlings was particularly tall, strong, and luxuriant, 
when the flower opened, it was single ; on the contrary, among the 
dwarf and stunted plants scarcely any but double ones could be seen. 
1 have found a great diiference in the proportion of double flowers 
from somfe sorts; but my present collection being composed entirely 
of the finest and newest varieties, the produce from the seed is much 
more regular. There is very little dependence on any producing a 
progeny of the colour of its parent. 1 have seen a yellow raised 
from a pink — a white from a purple — a purple fi-om a yellow, and 
direct opposite changes many times. I have understood that the 
reputed best flower known, was raised from a semi -double, and the 
indefatigable gardener who raised it, had not grown one of the least 
merit for seven years previous. Three or four of the finest flowers 
that came out last year, were raised from a flower which produces 
numbers of semi-double and crippled bloom, and is of no value to a 
connoiseur. 
The method of training is, I think, the most important, and perhaps 
the most difficult of any part of the management. An espalier fence 
is certainly a secure and excellent plan; but if a mass is growing 
together in rows, they will draw each other up out of all bounds, and 
only show flowers on their tops. Single stakes are a very poor 
protection : I have seen some excellent collections destroyed by one 
gale of wind, when well secured on this plan. I give the preference 
to four or five stakes to each plant, to any other mode of training ; 
but to accomplish this properly, care must be taken to leave sufficient 
space between the rows, as well as between plant and plant, which 
must in some measure depend on the grower, and the kind of flower. 
To have convenient room to pass between them, will require from 4 to 
5 feet from row to row, and 5 to 6 feet or more from plant to plant. 
As the leading shoot advances, a strong stake should be placed to 
that. The laterals should all be trimmed oflF to a certain height from 
the ground, in proportion to the growth of the plant ; — say for dwarfs 
