BAEOMETER. 



117 



times when no action is required. In the appa- 

 ratus, one pole of the battery is in connection by 

 its wire, which passes through a brass cap, c, 

 above, with the mass of the mercury in the 

 shorter limb of a siphon barometer. From the 

 other pole proceed two separate conducting 

 wires, having in their course two electro-magnets, 

 m and TO', and terminating in platinum points, 

 p and p' : these points stand in quite or nearly 

 the same vertical line beneath the ivory (insu- 

 lating) block which supports them, and are 

 directed toward each other. A thin metallic 

 disc, d, is upheld and steadied by an arrange- 

 ment of delicate wires, one of which, rising 

 from the float, also penetrates through the lat- 

 ter and to a slight distance into the mercury 

 beneath it: the disc, so supported, stands be- 

 tween the two platinum points, but not neces- 

 sarily touching either of them. The size of the 

 float is such that, the support rising from it 

 being properly guided, the rise and fall of the 

 former with the mercury occurs without fric- 

 tion or other appreciable resistance. "When the 

 disc, d, stands midway between the platinum 

 points, so as to touch neither of them, the cir- 

 cuit is broken ; when it comes in contact with 

 either point, the circuit is completed through 

 the corresponding wire. 



The disc being midway between the platinum 

 points, its surfaces are distant from them re- 

 spectively a little less than the y^th of an 

 inch. The ivory block carrying these points is 

 movable, being at the end of a horizontal arm 

 that is raised or lowered by the turning of an 

 upright screw, S. This screw distance of 

 threads ^th inch is below inserted into a 

 horizontal wheel, IF, having 40 teeth: thus, 

 a movement of this wheel to the extent of one 

 tooth raises or lowers the screw, and so the 

 platinum points, the ^Vtfth of an inch. At 

 the opposite sides of the wheel are fixed small 

 wheels, a and a', each having one tooth. By 

 means of clock-weights these wheels are con- 

 tinually impelled in opposite directions ; but 

 their motion is prevented, so long as the mer- 

 cury column is at rest, by the position of the 

 armatures of the two electro-magnets standing 

 close to them, each of these by means of a 

 "detent" arrangement locking its respective 

 wheel. It will be remembered, of course, that 

 the" fall of mercury in the shorter limb of the 

 siphon barometer indicates an effective rise to 

 twice the amount in the longer column, thus 

 showing increase of pressure, of which the lat- 

 ter is the true measure ; and vice versa. 



Suppose now that, the mercury having been 

 at rest, increasing atmospheric pressure begins 

 to depress the column in the shorter limb of 

 the siphon : the float & sinks, bringing the disc 

 d against the lower platinum point p', and 

 closing the circuit through the electro-magnet 

 m' ; the latter attracts its armature, and re- 

 leases the wheel a', which makes a revolution, 

 carrying the wheel TF backward one tooth, thus 

 lowering the screw the ^g-jfth of an inch, and 

 the joint p' the same distance below the po- 



sition at which contact with the disc occurred; 

 and so long as the mercury and the float con- 

 tinue to go down, each g( >Voth of an inch 're- 

 peats the process, the platinum point continually 

 moving down in advance of the disc. Let, on 

 the contrary, a diminishing atmospheric press- 

 ure allow rise of the mercury and float: the 

 disc d is then elevated into contact with the 

 upper platinum point p, thus closing the cir- 

 cuit through m, bringing down its armature, 

 and unlocking the wheel a, by the revolution 

 of which the large wheel is moved forward one 

 tooth, and the screw, with the arm and the 

 point jo, is carried upward the ^nnr* n f an 

 inch ; and this process also is repeated as often 

 as the increasing pressure elevates the disc 

 again into contact with the upper point p ; or, 

 the ascending and descending movements alter- 

 nate, according to circumstances. 



A movement of ascent and descent in the 

 mechanism, corresponding to that of the mer- 

 cury, has now been secured ; each y^jth inch of 

 this, however, as already indicated, denotes a 

 change twice as great the -r^th of an inch 

 in the effective height of the mercurial column, 

 and so in the value of the atmospheric pressure. 

 In order, now, to trace the curve corresponding 

 to these varying pressures, the wheel TF, at the 

 same time that it moves the screw S, is by a 

 train of wheel-work made to impart a propor- 

 tionate movement to a second screw >S", an arm 

 and pencil from the latter tracing the pressure 

 curve on a sheet wound upon a cylinder made 

 to revolve regularly by clock-work ; the ratios 

 of parts in the connection are such that the 

 curve, as produced, is " magnified a little more 

 than three times [3.077 times, it appears] the 

 barometrical pressure." 



For obtaining the printed numerical record 

 of pressure, there is introduced each day into 

 the mechanism a record sheet marked' at the 

 proper intervals with the hours from (noon) 

 to 24, or again. The machine prints the frac- 

 tional parts only of the inch in height, and to 

 three places ; and, at the end of each 24 hours, 

 the observer notes upon the sheet the date and 

 the integer inches in height of the column. Of 

 three wheels which receive independent mo- 

 tions about the same vertical axis, each has set 

 in its periphery types for the numbers 1, 2, &c., 

 to 9, 0. These wheels give in order the thou- 

 sandths, hundredths, and tenths of an inch each 

 of the first two, of course, revolving 10 times for 

 once of the wheel next succeeding it ; and the 

 uppermost advancing one type (-^fh circumfer- 

 ence) for a movement of the screw <S through 

 Wjnrth o f an inch. The combination of wheels 

 is such as to " carry for ten " either forwards or 

 backwards. At the end, in this instrument, of 

 each hour, the proper types on the three wheels 

 are brought into line with each other and with 

 the proper hour mark on the record sheet ; and 

 at this moment, the same clock-work which 

 governs the cylinder receiving the pressure- 

 curve, and also carries forward the record sheet, 

 causes a small hammer to strike a cushion stand- 



