ARTIFICIAL ROCK WORK. 65 



believe is still, the demand for them, that any one ac- 

 quainted with the subject will know that the Alps, the 

 Appenines, and every mountainous chain in the moderate 

 climates have been ransacked for the purpose of adorning 

 these faint imitations of nature's stupendous piles. 



The first and great care in erecting rock work is to see 

 that it does not resemble a pile of loose stones, the next 

 that it is not built in a regular form, such as the segment 

 of a circle or right line, as I have seen recommended in 

 some works — then that the fragments of rock be of widely 

 different sizes — for instance, a few small stones may fill a 

 large interval between heavy masses, but there must 

 neither be a mass of immense blocks together, nor a num- 

 ber of small ones piled on each other. It is by no means 

 requisite that the whole rock work should constitute one 

 mass ; on the contrary, more variety is produced by hav- 

 ing it in separate masses, with passages occasionally nar- 

 row and ruggedly rising, so that it is necessary to climb 

 over a slight impediment to make the circuit — some 

 art is required in arranging the crevices, so that the soil 

 fit for each plant be not washed out by heavy rain, and 

 the roots laid bare ; the moss which grows on the surface 

 of barren rocks is excellent for filling the lower part of 

 these interstices, and in cases where plants that love a 

 damp soil are cultivated, a garden pot with the hole stop- 

 ped to hold water, and another with the plant placed in 

 it, may be easily concealed — where there is water which 

 might be made to trickle over the rock work this aid is 

 ^not required. Due attention must also be paid to the 

 aspect. Some flowers only open in the sunshine, others 

 are only half hardy, for these the south and sheltered side 

 is appropriate; ferns and many others, love the shade, and 

 Avill not support the parching rays of the sun, these may 

 clothe the northern aspect. 



I have already made the remark in a former communica- 



