91 SCIENCE OF GARDENING. PART 11. 



1489. Kewley's alarum-thermometer (Jig. 219.) consists of a glass tube (a, a), about ten 

 inches in length, hermetically sealed atone end, and united at the other to a capillary tube 

 (6, 6), with an intervening and also a terminating ball (c and </). Imagine this double tube 

 placed in a horizontal position, the largest tube, and half the intervening ball, filled with 

 spirits of wine ; and the smaller tube and half of both of the balls, with mercury. If the 

 tube is now fixed by its centre in a brass frame (e), and nicely balanced, it is evident that 

 every change in the temperature of the atmosphere will produce a change in the position 

 of the centre of gravity of the tubes. One degree of heat, by expanding the spirit, will 

 press on the mercury in the intervening ball (c), and drive part of it over to the termi- 

 nating tube (d), which end will, in consequence, descend like the beam of a pair of scales 

 or of a steam-engine. Hence a moving power of great nicety and certainty is obtained, 

 the details for the application of which, to the ringing of a bell at any distance, commu- 

 nicating by a wire (/ ), need not be here entered into. Suffice it to say, that by means of 

 a scale (), it may be set to any required temperature, and will give the alarm at a dif- 

 ference of even the fourth of a degree, either of depression or elevation. It may be oc- 

 casionally used in gardening, to convey some idea of the changes taking place in the 

 temperature of particular hot-houses, to the head gardener's room, in the night-time ; but 

 its most important uses are in domestic economy, hospitals, &c. This balance-thermo- 

 meter, as it may be called, has been also applied, by its ingenious inventor, to the open- 

 ing and shutting of windows or sashes, valves of chimneys, or flues, and steam-cocks, 

 and either to all of these purposes at once, or to any one of them. 



1490. Kewleifs regulating thermometer, or automaton gardener (fig* 217.), consists of a 

 particular application of the alarum thermometer just described. For this purpose, the 

 thermometer is made from two to three feet in length, and the same principle may be ex- 

 tended to any length, as ten or twelve feet, with a proportionate increase in the diameter. 

 The apparatus wliich Kewley applies to the thermometer, and which enables him to get 

 the power requisite for opening the sashes or windows of hot-houses or buildings of any 

 magnitude, is a metal cylinder (/j), generally of rolled copper, as being cheapest, from 

 seven to fourteen inches in diameter, and from eighteen inches to two feet in length, 

 with an accurately fitted piston (t). This cylinder is placed either within or without the 

 hot-house or room in any convenient situation, and a cistern, or a barrel of ordinary dimen- 

 sions, filled with water, is placed on an elevated situation, say on a level with the chimney- 

 tops. The deeper the cylinder is sunk, the less the cistern requires to be raised above 

 the level of the floor of the house. If, as is often the case, a pipe of water is conducted 

 through the house from a distant reservoir of ordinary elevation, then nothing more is 

 necessary than attaching a branch-pipe. It is requisite that this pipe pass directly to the 

 point where the thermometer is placed, and at any convenient distance under it, not 

 higher than the bottom of the cylinder. Here it is joined to a tripartite cock (/t), whence 

 proceed two other pipes, one (l] to the cylinder, and the other (m) to a waste drain. The 

 stopper to this cock turns only to the extent of about one-fifth of a circle ; and when 

 turned to this extent to the right, it opens a communication between the supply-pipe (n), 

 and the cylinder (/*), when the pressure of the water in the reservoir, whether a barrel on 

 the top of a house or a distant cistern, raises the piston, and by a communication of cords 

 and pulleys with the sashes (o), they will be raised or opened ; and by another chain (;>), 

 the fire or steam-damper (<?), will be opened also. When the cock is turned to the left, 

 this communication is stopped, and one opened between the cylinder and waste-pipe (m), 

 by which the water escaping, the piston descends, and the sashes and dampers are shut. 

 The equilibrium of the balance-thermometer restored by the temperature, being reduced 

 or elevated to the proper degree, the plug is neither turned to the right nor left, and 

 every communication is closed. The cock is worked by two wires (r r), fastened to two 

 short levers, fixed on each side of the thermometer-frame, and the other ends of the cross 

 or handle of the cock (s s). To set the machine at work, it is only necessary to place the 

 scale to a degree at which it is desirable air should be given, taking care that the cistern 

 is not without water. A small cask of water, regularly supplied, will answer as well as 

 a large cistern, as the power is not as the body of water, but as its height. As a hot- 

 house seldom remains many minutes at the same degree of heat in the day-time, it is 

 evident that the sashes would be in almost continual motion, which, in houses where the 

 sashes open outwards, and especially the polyprosojnc, to be afterwards described, would 

 have a singular and animated effect in a flower-garden, or on a lawn. Where light 

 valves or ventilators are used, the balance-thermometer of this size has sufficient power 

 to open them without the aid of machinery ; and by lengthening the tube, sufficient 

 power may be obtained to open balanced windows in dwelling-houses, churches, or hos- 

 pitals. This machine was originally contrived for the use of the inventor's own garden 

 in Douglas (Isle of Man), and successfully employed to give air to pits and frames there 

 for two seasons. Having come to London, he employed it with the addition of more 

 machinery (see the patent, 1816) than he now uses, to ventilate a part of a house in the 

 New Kent Road, from 1816 to 1817. In 1818 he greatly simplified it, and thus im- 



