A FEW NOTES ON ROCK-GARDENS. 
451 
rockery or rock-garden. Such subjects, if practicable, should be 
planted by the square foot or by the yard, and a cheap supply 
would doubtless soon be forthcoming if only (in the language 
of economists) effective demand " for it arose. 
Thus far, I have been naming some of the best evergreen 
carpeters, useful, therefore, for the rockery's "framework." 
But I need scarcely say how much the list of carpeters will be 
increased when there come to be included in it those whose 
beauty mainly lies in their flowers, such as the Aubrietias, 
Alyssums, American Phlox, and a host of others. A good stock 
of large pieces of such subjects seems to be perhaps the most 
essential acquisition of any who would start the culture of rock- 
plants, if only in a border. The following are the evident 
advantages of starting with these large " pieces," rather than 
with small " year-old " plants. They are, so to speak, bound to 
grow, unless wilfully killed (an experience some of us have had). 
They rapidly yield further stock, and are therefore in the long 
run cheap. Only by means of them can you get large effect 
rapidly. Being surface-rooting they can be moved about as 
wanted from place to place during eight or nine months out of 
the twelve. 
Evergreen shrubs, then, dwarf only and generally choice, and 
carpeters " in masses, generally evergreen, are, in my view, 
the first and most important matter to give attention to in the 
furnishing of the rock-garden. 
I now pass on to make a few observations on the subject 
of the prevailing styles of rock-work structure. 
At the head of the short list I think must perhaps come the 
best work of Messrs. Pulham, as seen at Battersea Park and 
elsewhere. As regards size and large effect, there is still nothing, 
I think, to rival it. 
Some of their work, as you know, is constructed (as is to be 
preferred) from natural stone. Where such a size in the blocks 
is wanted as cannot well be had (or moved) in natural stone, 
clever imitation of it (in cement and otherwise) is resorted to, 
and those who are not in the secret would rarely detect imitation. 
It may be admitted that this artificial work is not yet perfect, 
and that in its earlier days it had many defects, now pretty well 
surmounted. It still has with many the bad character of being 
unsuited to receive plants. This character is now, I think, not 
