82 
.TOUPxNAL OF THE EOYAL IK JKTICULTlTxAL SOCIETY. 
suggest themselves as the work proceeds, but the principal effects should 
not only be planned beforehand, but it will also be found advantageous 
to at least roughly construct the parts intended for special effects at the 
beginning of the work and fill in the details afterwards, much in the 
same way as an artist wo aid proceed to paint a picture of a landscape. 
No plan can be worse than commencing to fix stones at one end and 
continuing to do so bit by bit in succession, till either the other end of 
the site is reached or the material at command is used up. 
Such continuous pieces of rockwork will always look more or less like 
a procession of stones, and can (in my humble opinion) never become 
things of beauty. 
No matter how large or how small the work is to be, it should never 
look like a collection of single stones, which is unavoidable if, as is so 
often done, the latter are placed at more or less regular intervals. If, on 
the other hand, the stones are combined together to form apparently 
groups of real rocks of various sizes, then the stones themselves will 
entirely lose their individuality, and this is as it should be. Even in 
the case of a single stone a great change might apparently be effected. 
Let us place, for instance, a stone of any size we like on a piece of 
green sward. It will look, what it is, a single stone, because we can 
plainly see the bottom edge. Now let us sink the same stone a few 
inches into the ground, and let us raise the turf a little so as to form a 
kind of undulating slope around it. What a transformation ! It looks 
like a stone no longer, but has assumed the appearance of the summit of 
a rock which has been upheaved from below till it pierced the surface, 
and might, for all the uninitiated would know, be continued underground 
for miles. 
Now let us place at some distance from this stone, not another single 
stone, but a group of ten, of fifty, or of a hundred stones, either joining 
each other, or connected apparently, by filling the interstices, not with 
cement or mortar, but with suitable plants. Let us fill the interior of the 
block thus formed with soil and stones, covering them with more plants. 
Let us further hide the bottom edges of all visible stones either by raising 
the turf around them, or by surrounding the base with a dense carpet of 
plants. Let us do all this with care, and the result will be almost magical 
in its effect. The stones will look like stones no longer, but will appear 
consolidated into a large block of real rock connected underground with 
the single stone previously referred to. Thus not stones, but apparently 
groups of rock, should form our rock-garden. 
These groups should, of course, vary in their shapes and sizes and 
in their respective distances from each other, being separated now by 
grassy banks, now by naturally grouped colonies of pla,nts, by rocky 
steps, or by a chasm, a cave, a pond, a streamlet, a piece of level sward, 
or any other apparently natural cause ; but they should never form an 
unbroken mass, not even in the smallest rockwork. 
In the case of igneous or unstratified rocks these groups should be 
as scattered and as varied as possible ; but in the case of stratified rocks 
the sedimentary character should be plainly visible in each group ; and 
this can be done Avithout actually placing the stones on each other, to 
the detriment of plant-growing. 
