38 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
This shifts the responsibihty of non- success in great measure 
from the plant to the planter ; and, to make my meaning clear, 
I will glance at one example. Take C. Baineri, a very dwarf 
Alpine with thickly pubescent herbage and somewhat thick 
underground stems. Its roots, to my mind, seem to have pecu- 
liar requirements both in regard to air and moisture, and their 
requirements may be met by keeping them near the surface, as 
on a moist stone ledge with a covering of well-consolidated soil, 
not more than 2 or 3 inches thick. With a deeper root-run I 
have found the roots to be rank in summer and to rot off 
in winter. On the stony ledge they seem to grow longer and 
faster, and to be much more wiry. We have, in fact, to humour 
a plant whose roots have very finely balanced requirements in 
the way of a freer atmosphere and moisture, whereas in regard 
to quality of soil it shows a corresponding amount of indiffer- 
ence. 
In the cultivation of Campanulas, whether in the style of a 
Campanula garden or otherwise, it would perhaps always be 
advisable to prevent the escape of seed, even if we sowed the 
seed otherwise, because when the seed is self-sown the unac- 
countable ways in which it comes up and develops will upset all 
our ideas of classification, and, worse still, the strongest and 
perhaps coarsest forms will obviously prevail, and in a very 
short time such self-sown seedlings will doubtless upset the 
best-arranged planting of carefully named varieties. So in- 
sidious do those seedlings seem to be that it almost appears as if 
they were actually conspiring to cheat you, especially when they 
implant themselves in the heart of a choice kind, or correctly 
named group, and in a comparatively short time might easily be 
taken by a casual observer to be the rightful owners of the 
positions and the names. 
in speaking of propagation I shall take the term in its more 
precise meaning, as the increase of a desirable variety by means 
of small parts taken from a common stock. I should hardly 
consider this analogous to "propagation with a spade," cutting 
off a slice of the matted roots of, say, such free growers as imsilla, 
car2)atica, or ktrhinaLa. The increase of these implies no care 
or skill to speak of, for they carry a full complement of all the 
parts essential to fr(;c growth, without artful or helpful means- 
I think it is rather important for the propagator to keep these 
