WALLS.] HORTICULTURE 221 dew. Any contrivance that serves to interrupt radiation, though it may not keep the temperature much above freez ing, will be found sufficient. Standard fruit trees must be left to take their chance ; and, indeed from the lateness of their flowering, they are generally more injured by blight, and by drenching rains, which wash away the pollen of the flowers, than by the direct effects of cold, i Hot walls, whether constructed to be warmed by flues or hot-water pipes, are nearly or quite obsolete. Their chief use is to assist in ripening the young wood and the crops of the later varieties of tender fruits by the aid of artificial heat, but the expenditure would be more usefully directed to the construction of a glazed house for that purpose. Espalier Rails. Subsidiary to walls as a means of training fruit trees, espalier rails were formerly much em ployed, and are still used in many gardens. In their simplest form, they are merely a row of slender stakes of larch or other wood driven into the ground, and connected by a slight rod or fillet at top. The use of iron rails has ,now been almost wholly discontinued on account of metal lic substances acting as powerful conductors of both heat and cold in equal extremes. Trees trained to espalier rails have some advantages, as they are easily got at for all cultural operations, space is saved, and the fruit, while freely exposed to sun and air, is tolerably secure against wind. They form, moreover, neat enclosures for the vegetable quarters, and, provided excess of growth from the centre is successfully grappled with, they are pro ductive in soils and situations which are suitable. 10. Plant Houses. These include all those structures which are more intimately associated with the growth of ornamental plants and flowers, and comprise conservatory, plant stove, greenhouse, and the subsidiary pits and frames. They should be so erected as to present the smallest extent of opaque surface consistent with stability. With this object in view, the early improvers of hot-house archi tecture substituted metal for wood in the construction of the roofs, and for the most part dispensed with back walls; but the conducting power of the metal caused a great irregularity of temperature, which it was found difficult to control ; and, notwithstanding the elegance of metallic houses, this circumstance, together with their greater cost, and some doubt as to their durability, has induced most recent authorities to give the preference to wood. The combination of the two, however, as in the Crystal Palace at Sydenham, shows clearly that, without much variation of heat or loss of light, any extent of space may be covered, and houses of any altitude constructed. The earliest notice we have of such structures is given in the Latin writers of the 1st century (Mart., Epiyr., viii. 14 and GS) ; the ASwvtSos /oyvroi, to which allusion is made by various Greek authors, have no claim to be mentioned in this connexiim. Columella (xi. 3, 51, 52) and Pliny (II. N., xix. 23) both refer to their use in Italy for the cultivation of the rarer and more delicate sorts of plants and trees. Seneca has given us a description of the applica tion of hot water for securing the necessary temperature. The botanist Jungermann had plant houses at Altdorf in Switzerland ; those of Loader, a London merchant, and the conservatory in the Apothecaries Botanic Garden at Chelsea, were the first structures of the kind erected in British gardens. These were, however, ill adapted for the growth of plants, as they consisted of little else than a huge chamber of masonry, having large windows in front, with the roof invariably opaque. The next step was taken when it became fashionable to have conservatories attached to mansions, instead of .having them in the pleasure grounds. This arrangement brought them within the province of architects, and for nearly a century utility and fitness for the cultivation of plants were sacrificed, as still is often the case, to the unity of architectural expression between the conservatory and the mansion. Plant houses must be as far as possible impervious to wet and cold air from the exterior, provision at the same time being made for ventilation, while the escape of warm air from the interior must also be under control. The most important part of the enclosing material is necessarily glass. But as the rays of light, even in passing through transparent glass, lose much of their energy, which is further weakened in proportign to the distance it has to travel, the nearer the plant can be placed to the glass the more per fectly will its functions be performed : hence the import ance of constructing the roofs at such an angle as will admit the most light, especially sunlight, at the time it is most required. Plants in glass houses require for their fullest development more solar light probably than even our best hot-houses transmit, certainly much more than is transmitted through the roofs of houses as generally con structed. 1 Plant houses should be constructed of the best Baltic pine timber, as being the most durable, but the whole of the parts should be kept as light as possible. In many houses, especially those where ornament is of no conse quence, the rafters are now omitted, or only used at wide intervals, somewhat stouter sash-bars being adopted, and FIG. 4. Lean-to Plant House. stout panes of glass, 8 to 12 inches wide, made use of. Such houses are very light ; being also very close, they require careful ventilation. The glass roof is commonly designed so as to form a uniform plane or slope from back to front in lean-to houses (fig. 4), and from centre to sides in span-roofed houses. In some cases, however, the roof sashes are fitted up on the ridge-and-furrow principle invented by Sir Joseph Paxton, shown in fig. 5, w-hich represents the original ridge-and-furrow house erected by him at Chatsworth. To secure the greatest possible influx of light, some scientific horticulturists recommend curvilinear roofs ; but the superiority of these is largely 1 Mr Knight, an unquestionable authority, proposed a general pitch or elevation of 34 for the latitude of London, the angle at which the rays of the mid-day sun are perpendicular to the surface on the 20th of May and 21st of July. This would afford four months, from the 20th of April to the 21st of August, during which the angle of incidence at mid-day would not at any time amount to 9, while the deviation at the winter solstice would be 43, and the loss of light from reflexion would be little more than ^ The an S le of 45 has becn recom mended as a pitch extremely suitable for early vineries and pine stoves, in which case the midsummer deviation would be lt, and the loss &> and the winter deviation 30, the loss being nearly the same. Greater exactness, however, has been sought in this matter than is at all neces sary. The reduction of the opacity of the roof arising from the breadth and depth of rafters and astragals is of much more consequence. The massive rafters, framed sashes, and inferior glass inserted in small fragments, with numerous overlaps liable to be choked with dirt, intercept a large proportion of the solar light and heat in ordinary
glass houses.
