K(^rK'-r,ARDENS. PONDS, AND STREAMLETS. 
83 
T ofton find it a much better plan, especially in the case of choice 
Alpine plants, to form more or less narrow vertical crevices, but keeping 
the tops of the stones so as to form irregular little terraces, appearing 
like strata. 
Neither is it necessary to take such elaborate pains to have the strata 
throughout the work running at the same angle. 
On the contrary I consider it a mistake to do so, and for the sake of 
greater variety prefer to have such groups of rock as are distinctly 
separated from each other, showing the strata at a different angle from 
the adjoining group — which would suggest that the separation and the 
scattering of the rocks was the result of volcanic action or other violent 
forces. Since in nature such scattered groups are not only found in 
abundance, but constitute the most picturesque scenery, we need have 
no hesitation in adapting this principle. 
The Planting of a Eock-gakden can only briefly be mentioned 
here ; but a full list of plants for various purposes and for different situa- 
tions will be published in due course as an appendix to this paper. I 
will therefore mention only a few facts in connection with plants and 
their requirements in the rock-garden. It is, I think, a great mistake to 
let one man erect the rocks and to let another man plant them when the 
stone-work is finished. "Whoever designed or superintended the stone- 
work should also be responsible for the planting ; otherwise intended 
effects may easily be lost ; for two men, however good they may be, can 
never have precisely the same ideas. 
A good plan is to plant some of the depressions and recesses with 
very dwarf stuff only, so as to emphasise the more x^rominent portions, 
which latter should be planted with taller things. But if planted by a 
novice the reverse would probably be done, and instead of the intended 
effect being emphasised it w^ould be utterly lost. Then, again, during 
the construction of the rocks an expert will prepare the soil according to 
the special nature and requirements of the plants he intends to use, and 
of whose effects he is able to judge beforehand. The choicer kinds of 
Alpine plants, for instance, require in most cases rock specially suited to 
them, such as narrow crevices filled partly with pieces of broken lime- 
stone, or (in the case of lime haters) with granite or sandstone mixed 
with the soil. These plants demand moreover greater watchfulness 
against the ravages of slugs and other unwelcome intruders, and it is 
best, therefore, to devote certain groups of rocks entirely to these moun- 
tain gems, planting the latter, not singly, but in groups or masses and 
choosing for their companions such as would either harmonise with them 
in colour, or would flower at a different season, and thus ensure succes- 
sion of bloom. 
But above all things the small and slow growing plants should be 
associated with such companions as would grow about at the same rate 
without the danger of overcrowding and killing each other. 
In the same manner we may have rocky groups for plants of quicker 
growth, for ferns, for rock-scrubs, &c., but in planting we must always 
look ahead and allow for the full development of the plants rather than 
be misled by immediate effect. 
Vertical or abruptly sloping fissures between the stones are, as a 
