U()CK-GA1M)KNS, roND.s. AND .S TJiKAM LJ-:'1'S. 
85 
The most pleasiiiij: form of water in a rock-garden is, I think, a spi ing 
emerging apparently from a cleft among the rocks and then winding its 
way through the rock-garden as a murmuring streamlet, forming a water- 
fall or two in its course and then flowing more lazily and broadening 
out into an irregular pond, the overHow of which might supply a bog-bed 
or two, or emerge again from the pond as a rippling brooklet, which is 
finally lost to view beneath a rock. 
All this, and more, might be done with only a comparatively small 
supply of water, and if well done there should be no visible trace of any 
artificial work, but the whole should look like the work of Nature. 
To accomplish this result we must learn a few lessons from Nature 
before we attempt to imitate her. 
In arranging for a waterfall, for instance, there is a strong tempta- 
tion (if the water is introduced by a pipe) to place this as high as 
possible, sometimes even in the highest part of the rockwork. This is 
wrong. Unless we can see beyond the waterfall still higher ground 
which apparently supplied its source, such a fall will look unnatural, and 
therefore ugly. In the case of a streamlet the fiow^ing water, which 
meets with an obstacle, say a rock, will be diverted at an angle, and, if 
the ground be soft, a hollow place will be washed out on the opposite 
bank. Hence the curves of our artificial streamlet must be arranged 
accordingly, and where a convex line occurs on one side the shore should 
be concave on the other, and vice versa. 
As ponds are in most cases simply broadened streamlets this rule 
applies to them also. 
In planting the sides of a pond the most natural result will be 
obtained by planting thickly the portions which project into the w^ater 
and leaving the concave parts comparatively bare. This will also have 
the effect of obscuring some parts of the pond from view when seen from 
any particular point, with the result that the pond will appear larger 
than it really is on account of the artificial perspective thus created. 
For securing the sides and bottom of ponds and streamlets there is 
no better material than cement-concrete. But this must be entirely 
hidden from view, namely, on the bottom by river-gravel and pebbles of 
all sizes, and on the sides by rocks and grassy banks. The latter should 
dip at least iS inches below the surface of the water, as otherwise the 
ripple of waves, however small, would undermine them. 
The roots of Iris, Spiraeas, Astilbe, and other waterside plants planted 
into such banks would have free access to the water, and would soon feel 
at home. 
For Nymphnsas and other aquatics I find the best way of preparing 
a home for them is to build up with loose stones various compartments 
and filling them with soil till a depth of 15 inches or so below the water- 
level is reached. Or in the case of large ponds the Water-lilies can be 
sunk in baskets to the same level. 
Artificial bog-beds, which might, as already indicated, be fed by the 
overflow of the pond, are also best secured by cement-concrete, and so 
arranged that supply and drainage can be regulated. Since the w^hole of 
the cement work would be covered with soil, &c., its ugliness does not 
matter. 
