BOOK III. 



HOT-HOUSE FURNACES AND FLUES. 



325 



1654. The sides of common flues are commonly built of bricks placed on edge^ and the 

 top covered by tiles, either of the full width of the flue outside measure, or one inch nar- 

 rower, and the angles filled up with mortar, which Nicol prefers, as neater. Where a 

 stone that will endure fire-heat without cracking is found to be not more expensive than 

 tiles, it is generally reckoned preferable, as offering fewer joints for the escape of the 

 smoke. Such stones are sometimes hollowed on the upper surface, in order to hold water 

 for the benefit of plants in pots, or for steaming the house. 



1655. Broad and deep flues, agreeably to the Dutch practice, have been 272 

 recommended by Stevenson (Caled. Mem.) ; that of making them narrow 



and deep, agreeably to the practice in Russia, is recommended by Oldacre, 



gardener to Sir Joseph Banks, and that of using thin bricks (Jig. 272.) 



with thick edges, by S. Gowen (Hort. Trans, iii. ) In Gowen's flues, the 



section (Jig. 273. a) shows less materials 



than any other brick flue, the covers (6) and 



the side wall bricks (c) being quite thin, 



the base requisite for building the latter on 



one another being obtained by the thickness 



of their edges (d, e), which is equal to that 



of common bricks. * 



1656. Can-flues (jig. 274.), long since 

 used by the Dutch, imbedded in sand, and 

 for the last fifty years occasionally in Eng- 

 land, are sometimes employed. They consist 



of earthen pipes, straight (a), or rounded at the ends for returns (b), and joined together 

 by cement, placed on bricks (c). They are rapidly heated, and as soon cooled. None of 

 the heat, however, which passes through them, can be said to be absorbed and lost in the 

 mass of enclosing matter, as Knight and Sir Joseph Banks (Hort. Trans.) assert to be the 



274 



275 



case with common flues. They are only adapted for moderate fires, but judiciously 

 chosen, may frequently be more suitable and profitable than common flues ; as, for 

 example, where there are only slight fires wanted occasionally ; or where there is a re- 

 gular system of watching the fires, in which case, but not otherwise, the temperature can 

 be regulated with sufficient certainty. 



1657. Ttie embrasure flue (flg.275.) is the 

 invention of Sir G. Mackenzie, and is by him 

 strongly recommended, as exposing a greater 

 heated surface in proportion to its length. 

 (Hort. Trans, vol. ii. p. 175.) 



1658. Cast-iron flues have also been recom- 

 mended on account of their durability, but unless they were to be imbedded in sand, or 

 masonry, they are liable, in an extreme degree, to the same objections as can-flues. A 

 triangular cast-iron flue, to be coated over with a mixture of one part clay and three of 

 sand, is recommended for trial by Sir G. Mackenzie. (Hort. Trans, v. 216.) For our 

 part we cannot perceive a single circumstance in favor of its adoption. 



1659. The best sort of flues, after all that has been said on the subject, is, in our opi- 

 nion, the common form, built of thin well burned bricks neatly jointed, with the bottom 

 and top of tiles, and no plaster used either inside or outside. Where only one course of 

 a flue can be admitted the broader it is the more heat will be given out as it proceeds, and 

 as a consequence, one extremity of the space to be heated will be hotter than the other ; 

 a return or double course of a narrow flue is, therefore, almost always preferable to one 

 course of a broad flue. With respect to the embrasure flue, flues with iron tubes, or iron 

 covers, and various others that have been recommended or described in recent volumes of 

 the Horticultural Society's Transactions, they are liable, in our opinion, to great objections, 

 and chiefly to produce sudden excesses of heat, and in general as tending to extremes of 

 temperature. 



1660. The size of flues is seldom less than nine inches wide, by fourteen or eighteen 

 inches high inside measure, which suits a furnace for good coal, whose floor or chamber is 

 two feet long, eighteen inches wide, and eighteen inches high. According as the object 

 varies, so must the proportion both of furnaces and flues. (Designs for Villas, &c. 1812; 



Y 3 



