KOCK-CAKDKNS, I'oNDS. AND .STKKA.MLKTS. 
81 
Stones of a f^hiring light colour should l)e avoided, and on no account 
should pieces be used which are covered with glassy crystals. Stones of 
the latter class may be pretty to look at individually, but, for being 
combined together to form picturesque effects in a rock-garden, they are 
utterly unsuited and give the work a downright vulgar appearance. The 
beauty of a rock-garden should depend on its natural arrangement 
generally, and not on individual stones. 
For constructing rocks, true to nature at least, a little geological 
knowledge is indispensable, (leologists divide all rocks into two great 
classes, namely, the unstratified, or igneous, and the stratified, or sedi- 
mentary. 
To the first-mentioned group belong the plutonic rocks, which through 
heat have been upheaved from the interior of the earth, such as granite 
or porphyry, and the volcanic rocks, like trap, greenstone, or basalt, which 
are the direct result of volcanic forces, and consist almost entirely of lava. 
The second class, namely, the stratified or sedimentary rocks, were 
formed under water, and include different kinds of limestone, slate, oolite, 
sandstone, flint, chalk, &c. 
Allied to these arc metamorphic rocks, which were crystallised through 
contact with heated rocks of the plutonic class. 
In constructing a rock-garden stratified and unstratified rocks should 
never be arranged in the same way, least of all when the work is on a 
large scale. If the stones at disposal are of the igneous kind it will be 
found advisable to select as many different shapes and sizes as possible ; 
but if they belong to the sedimentary class flat stones and long narrow 
stones of all sizes should have the preference. 
For small rockworks stones varying in weight, say from I cwt. 
to 15 cwt., will prohably be found large enough to manage ; but when 
the work is on a larger scale, and facilities exist for using proper appliances 
for lifting, some pieces w^eighing even several tons may be used wdth 
advantage ; but in any case let us have as much variety as possible. 
The General Arrangement of a rock-garden is naturally of the 
utmost importance. Not only is this almost entirely a matter of taste 
for which hard-and-fast rules are out of the question, but in addition to 
this difficulty, if such it can be called, there are generally a host of other 
things to be considered, such as the purse of the owner, the climate, the 
amount of care required in looking after the rock-garden after its 
completion, the size of the work, the material at disposal, and several 
other items. 
After due consideration of all these points the work is best commenced 
by marking it out in the ordinary way with pegs. If a pond or streamlet 
be contemplated this should be marked out first, as the soil gained 
through its excavation would probably be required in the higher parts 
of the work, which should also be at least roughly marked out. In order 
to judge of the various effects in planning the work I have always found 
it a good plan to let several workmen hold up at the same time pieces of 
wood or other materials at the different heights and angles, wdiich the 
various groups of rock would eventually occupy, so as to have at least 
an approximate idea before starting of the ultimate effects. 
No doubt slight alterations and probably further improvements may 
