116 



BAKOMETER. 



tion of the tube made with the upright pipe. 

 Now, opening all the taps, the water displaces 

 the air both ways from the barometer tube, 

 until it flows freely from the lower end of this : 

 the tap D is then closed, and the end of the 

 tube temporarily stopped ; and the cistern, pre- 

 viously filled with olive oil, is brought up into 

 place, so as to receive the end of the tube, the 

 latter having been meanwhile again opened. 

 The oil is then drawn off, until about three 

 inches' depth of it remain to protect from air 

 the water, which meantime flows in under- 

 neath it. Then closing 2?, the other three taps 

 being now open, the water fills the upright tube 

 until it runs from the upper end : A and are 

 then closed, and B opened ; and the water de- 

 scends by gravity until its height is such that 

 it balances the pressure of the atmosphere on 

 the liquid surface in the cistern. The level of 

 the surface of oil in the cistern being brought 

 to correspond with the height of the marked 

 zero point, an observation of the height of the 

 water-column can then be taken : when the 

 mercurial barometer showed 30.4 inches, the 

 temperature being 67, the water-column had a 

 height of 400 inches. The estimated cost of 

 the instrument, exclusive of the gas-fitter's 

 time, was about $15. 



In the Philosophical Transactions of the 

 Royal Society, for 1832, is found a description 

 by Mr. Daniell of a water-barometer, erected 

 by him at the Society's rooms, Somerset House. 

 He stated that, in this, " the water appears to 

 be in perpetual motion, resembling the slow 

 action of respiration." This Mr. Bird confirms : 

 the oscillations occur about every four, hours 

 and twenty seconds; and as they ordinarily 

 vary from but about ^ to^V of an inch, they re- 

 quire to be watched with a magnifier. Mr. 

 Bird finds that the most surprising oscillations 

 take place during a thunder-storm accompanied 

 with great falls of hail and heavy rain-drops. 

 He gives a chart, with curves of the water- 

 barometer and the mercurial, during one hour 

 and five minutes of such a storm, July 20, 

 1859. Comparison of the two curves shows 

 that the water-column responded both sooner, 

 and also more fully, than the mercurial the 

 whole curve of the former running higher, its 

 oscillations beginning earlier and standing higher 

 respectively. After heavy thunder, and a heavy 

 fall of rain and hail, as the sky commenced to 

 brighten, the water rose very suddenly in five 

 minutes more than -^ths of an inch. The au- 

 thor thinks that, in accordance with Sir J. 

 Herschel's theory, that the electric discharge 

 is the result and not the cause of the sudden 

 condensation of cloud into rain, the sudden in- 

 crease of pressure of the air may be due to the 

 speedy and great precipitation of vapor, leaving 

 the air mo^e dry and dense. 



Barographs, or Self- Register ing Barometers. 

 The tendency of the past few years to auto- 

 matic instruments for purposes of scientific ob- 

 servation, began quite early to be developed in 

 connection with the barometer. The earliest 



properly self-registering instrument of this class 

 appears to have been that of Mr. Alexander 

 Keith, of Scotland: the essential parts of this, 

 in its second and more complete form, were, 

 the siphon barometer, and a float on the sur- 

 face of the mercury in Us shorter limb, this 

 carrying an upright wire and a pencil which 

 traced, on a sheet ruled for a month's record 

 and slowly moved by clock-work, a line cor- 

 responding to the varying heights of the mer- 

 cury. Mr. Bryson, of Edinburgh, later sub- 

 stituted a minute knife-edge, which, at the end 

 of every hour, a hammer forced into the pre- 

 pared surface of a daily record sheet rolled upon 

 a cylinder ; by this means he avoided the inac- 

 curacy due to friction of the pencil. Previous 

 to 1858, the photographic registering of barom- 

 etric variations was in operation at the Royal 

 Observatory, Greenwich : in this, the light of 

 a lamp being continually directed through the 

 Torricellian vacuum of a standard barometer, the 

 lower margin of the trace which it produces on 

 a slowly advanced sheet of sensitive paper, 

 forms the barometric diagram or curve. Pre- 

 vious to 1859, also, M. Hardy, of Paris, applied 

 electricity to the work of registering : he used 

 the siphon, and the float with its wire and pen- 

 cil ; while at the end of every five minutes, the 

 current, started by the movement of clock- 

 work, caused a slight blow to be given to the 

 tube, to bring the mercury to its true level, and 

 a half minute later a second current caused a 

 pressure to be applied to the pencil, so as to 

 mark its position on the record sheet. 



Hough's Typo-Barograph. Professor G. "W. 

 Hough, Director of the Dudley Observatory, Al- 

 bany, has recently invented a highly complete 

 method of automatic registering, in an instru- 

 ment termed by him in his descriptive pamphlet 

 (from which this account is drawn), an " Auto- 

 matic Registering and Printing Barometer," and 

 later named the Typo-Barograph. He has aimed 

 to furnish a process simpler than the photograph- 

 ic, and susceptible also of quite general applica- 

 tion. "The problem to be solved," he says, 

 " was to cause any meteorological instrument, 

 by means of suitable mechanism, simply and ef- 

 fectually to record its own changes." To this 

 end it was among other essentials necessary 

 that the mechanism should be such as to repeat 

 the changes within the instrument in all their 

 forms of motion uniform or variable, forward 

 or reverse. 



Beginning with the barometer, it was found 

 by experiment that, by employing the float 

 and making connection in the electric circuit 

 outside the tube, a motion of less than .0005 

 = 2tWth of 'an inch could be readily shown. 

 It was desirable that the registering apparatus 

 should not only produce a linear diagram or 

 curve of the varying atmospheric pressure, but, 

 in order to secure exact results, that it should 

 also furnish simultaneously a printed numerical 

 record' of the same. A Daniell's battery is 

 employed, and the " make " circuit adopted, as 

 saving consumption of the battery elements at 



