I I I !
partial Kiwtli of tlio trees, or oven bm-ning those parts of thorn immediately opposite the
fni-iiacc To provont accidents of this kind, the fiu-nace mnst always bo placed at some
distance, say from 18 in. to 3 ft. from the hack of the -»vail. An unprovomont haa been
made in flu? walls by Mr. ShieUs, of Ersldno House, Eenlrcwshn-c, by the introduction
of a register, to admit, at pleasure, a part of the smoke and heat from tlie fiic diicctly
into the second comsc of flues. ^ • .1. n 1 •i*.
2020. T he cellular w a ll (Jig . 581.) differs from the ilued wall, m having the wall built
hollow, or at least with communicating vacuities equally distributed from the surface of
the o-round to the coping. (H o rt. T ra n s ., vol. iv.) If the height does not exceed 10 or
12 feet, these walls may be formed of bricks set on edge, each course or layer consisting
of an alternate series of two bricks set edgewise, and one set across forming a thickness
of 9 in., and a series of cells, 9 in. in the length of the wall, by 3 m. broad. Ihe
second course is laid in the same way, but the bricks alteraate or break joint with
the first. The advantages of this wall are obviously considerable m the saving oi material
and in the simple and efficacious mode of heating; but the bncks and mortax
must be of the best quality. This wall has been tried in several places near Chichester,
and at Twickenham, by F. G. Carmichael: it is found to succeed perfectly as a h^ot-
wall, and at 10 ft. high to be sufficiently strong as a common garden-wall,^ with a
saving of one brick in three. As a whole, indeed, it is stronger than a solid nme-inch
wall, on the same principle that a hollow tube is less flexible than a solid one. It is
evident, that the same general plan might be adopted in foi-mmg cellular walls ol
gi-eater height, by increasing their width. A veiy high waU might have two ^'Stems ot
cells divided vertically, one or both of which might be heated at pleasure. The same
idea may be advantageously appHed to flues for heating hothouses by steam, and for
other purposes. Piers may be formed either on both sides of the wall (a ), or on one
side by bricks on edge (b ), so as.to bond in with the rest of the work.
2021. H o llo w w a lls may also be formed by using English instead of Flemish bond ;
that is, laying one course of bricks along each face of the wall on edge, and then bonding
them by a com-se laid across and flat. Such a mode has been practised and described
by Dearne, an architect in Kent. , , „ n 1 n j i
2022 W here w a ll- fr u it is an object o f consideration, the whole of the walls should be
flued 01- cellular, in order that in any wet or cold autumn, the fr-uit and wood may be
ripened by the application of gentle fires, night and day, m the month of September.
It is an error to light the fires of hot-walls only in the evenings, the effect ot heat in the
process of maturation being much greater when accompamed by light. In all hot-walls
one precaution must not be neglected, the building in, on the inferior or outer side,
small cast-iron doors, or framed stones, which may be opened at pleasure, in order to
withdraw the soot. They must be made perfectly air-tight, which is readily accomplished
by having double cast-iron doors, in what is called Count Rumford s manner.
2023. T he m ud o r e a rth -w a ll is formed of clay, or better of brick earth 111 a state between
moist and di-y, compactly rammed and pressed together between two movable
boarded sides retained in their position by a frame of timber, which form between them
the section of the wall; these boarded sides are placed inclining to each other so as_ to
form a wall tapermg as it ascends. One layer of the length of 12 or 20 feet being
completed, another layer is formed on that, and so on. till the wall attains the given
height, which in the Netherlands, and some parts of Gei-maiiy, where these walls prevail,
seldom is above 10 ft. A succinct account of the mode of forming these walls
in Devonshii-e wiU be found in our E nc yc. o f Cottage, F a rm , and V illa A rchitecture,
§ 839.
2024. B o a rd ed o r wooden w alls (Jig . 582. d ) arc vai-iously constnicted. One general
rule is, that the boards of whicli they are composed should either be imbricated or close •
jointed, in order to prevent a current of air
from passing tlu-ough the seams; and in
either case they should be well nailed to
the battens behind, in order to prevent them
from wai-ping from the sun. When well
tan-ed aud afterwards pitched, such walls
may last many years. They must be set on
stone posts, or the main parts or supports
formed of cast iron. Nicol informs us
(K a le n d a r, p. 149.) that he has “ con- _ _
structed many hundred hiical feet of wooden walls, which recline considerably towards
the north (Jig . 582. b), presenting a surface at a better angle with the sun than if they
were upright. They are placed on sloping ground, and range in five ranges or lines,
due east and west, at the distance of 7 yards from each other, the southmost being
5 ft. high, and the northmost 7 ft., composed of imbricated boards, pitched over to
give them durability; the supports are set on (not in) blocks of stone, which are sunk in
the earth, and firmly laid 011 solid foundations, 3 ft. under the ground level.”
2025. T em p orary w a lls and screens may be formed of reeds, rushes, or straw. Wliat
is called a reed wall (Jig . 583. aud 584.) may be described as 10 ft. high, and consisting
of a double trellis (a b), composed of horizontal laths
about 8 in. apart, and a coping board (c) 9 in. broad ; the
583
reeds are placed endwise within the trellis (d ), and supported
about a foot from the ground, to keep them from
rotting ; this interval of a foot being filled up witb slates,
placed on edge (e). The trellis rods are nailed to posts
(Jig . 584.), and, by taldng off a few of these rods on one
side, the reed mats can be removed and renewed. Russian
mats would, no doubt, answer very well, and last a long
time, and they might be taken out with still less trouble.
Straw mats would also do, where reeds could not be got;
and heath, as being of a dark colour and very durable,
would make the best of all structures of this kind.
Peaches, grapes, and other fruits, ripen just as well on
these structm-es as on brick walls.
2026. In c lin e d fr u it-w a lls seem to have been first suggested,
about the beginning of the eighteenth centm-y, by
N. F. De Douillier, F. R. S., an able mathematician, author
of a work, entitled F ru it-w a lls improved by inc lin ing them
to the H o rizo n, &c. Some walls were formed at Belvoir
Castle oil this plan, which Switzer informs us he went
to see, but found them damp, and the trees liable to be
injured by perpendicular fr-osts. De DouiUier’s work, as
being the production of a speculative theorist (he was
tutor to the Marquess of Tavistock), appears to have been
rejected by Miller, Switzer, Laivrence, and the designers
of gardens of that day, but it is replete with ingenuity
and mathematical demonstration, and wcU illustrates the
importance of sloping walls where they are to be protected
by glass or gauze. In a communication to the
Horticultural Society (vol. iv. p. 140.), by Stoffels, gardener
at Mechlin, he states, “that he had an opportunity
of comparing the effect of a sloping and pci-pen-
dicular wall in the same garden, for the growth of peach
trees, and that the result was greatly in favom- of the
foi-mcr.”
2027. T h e w a vy o r serpentine w a ll ( f g . 585.) has two avowed objects ; first, the saving
of bricks, as a wall in which the centres of the segments composing the hue are 15 ft,
apart may be safely carried 15 ft. high, and only 9 in. in thickness from the foundations
; and a four-inch wall may be built 7 ft. feet high on the same plan. The next
proposed advantage is sbelter from all winds in the dii-ection of the wall; but this ad-,
vantage seems generally denied bv practical men.
r p 3