THE BAROWETEK.] 



METEOROLOGY. 



1125 



of which is attached to a small weight, N. From this 

 arrangement it will be seen that every rise and fall of 

 the real barometric column, in the long arm of the tube, 

 will correspond with a parallel fall and rise of the mer- 

 curial column in the short arm. It will be seen, moreover, 

 how the small float, W, is raised and lowered ; how the 

 pulley, P, will be caused to revolve, and the index- 

 hand to traverse the dial-plate of the instrument. The 

 exterior of the wheel barometer is represented in 

 Fig. 10. 



We need scarcely state that the wheel barometer is 

 considered, pneumatically, a very 

 imperfect instrument. Not only 

 is the varying ratio between the 

 mercurial column and the level 

 of the mercury in the reservoir 

 here a necessity the very index 

 motion depending upon it but 

 the presence of the float, W, 

 tends also to embarrass the free 

 ascent and descent of the baro- 

 metric column of mercury. 



Manufacture of a Correct Baro- 

 meter. Not to render the prin- 

 ciples concerned in the barometer 

 complex, we have hitherto as- 

 sumed that the act of charging a 

 tube with mercury is simple and 

 free from ditticulties. Practically 

 this is not BO ; many precautions 

 have to be taken, otherwise the 

 resulting barometer will be any- 

 thing but correct. The tube 

 elected must not be too small ; 

 many instruments are rendered 

 incorrect owing to neglect of this 

 precaution. The internal dia- 

 meter of the barometer tube 

 should never be less than a quar- 

 ter of an inch ; it may be even 

 more with advantage. A small 

 barometer tube prejudices the 

 correctness of the instrument in 

 two respects, for the variations 

 of expansion and contraction of 

 the mercury, due to variations of temperature, are 

 more considerable ; and the motion of the quick- 

 silver, up and down, is impeded by friction against the 

 glass. 



The Mercury miwt be Pure, and deprived of Atmospheric 

 Air. Mercury, as commonly existing, in generally im- 

 pure. It contains uncertain quantities of tin, lead, and 

 sometimes zinc. Of cour.-o these admixtures damage 

 the mercury for barometric purposes. The observer 

 desires to read off his atmospheric pressure in terms of 

 inches of mercury, not in terms of inches of a mercurial 

 compound. It is indispensable, therefore, that the im- 

 purities be discharged or extracted. Various processes 

 are used to this end; but that usually followed by makers 

 of barometers, consists in agitating the mercury to be 

 purified, with dilute nitric acid, which gradually dissolves 

 out the extraneous metal, and leaves the mercury 

 pure. 



Far greater difficulties arc encountered in discharging 

 atmospheric air from the mercury employed. This is 

 accomplished by boiling the mercury after it has been 

 poured into the tube. The operation requires great 

 delicacy, and the instrument is frequently broken in the 

 operation. 



Method of Reculinq off Barometric Indications correct!]/. 

 It is not possible to read off, by referring to a common 

 scale of inches and parts of inches, the various small 

 elevations and depressions of a barometric column. 

 Two methods are had recourse to for obviating this 

 difficulty : one is the diagonal tube, a contrivance 

 altogether peculiar to the barometer ; the other is the 

 Vernier or Nonius scale, employed for the general 

 purposes of facilitating the reading of minute scale 

 divuiona. 



Fig. 11. 



The diagonal-tubed barometer is represented by the 

 accompanying diagram (Fig. lift. 12. 

 11). ^^. 



Tlie Nonius or Vernier Scale. 

 This is an ingenious contri- 

 vance for measuring small 

 linear divisions by means of 

 larger divisions, and, conse- 

 quently, more easily recognisa- 

 ble by the eye than larger 

 divisions would be. Thus, for 

 example, by means of a Vernier 

 graduated in divisions of an 

 inch and one-ninth, we can read 

 off tenths of an inch, as will 

 be seen by reference to Fig. 12. 

 Let A be a scale sliding in 

 proximity to B. Let each of 

 the divisions on B be = one 

 inch, and each of the divisions on A = one inch 

 and one-ninth. From these considerations it 

 follows that nine divisions on A are equal to 

 ten divisions on B. Directing the eye to the 

 upper limit of A, it will be seen that its edge 

 corresponds to thirty inches, and tometliini 

 more on B. In this case the observer would 

 have no difficulty in recognising the amount 

 over and above thirty inches on B to be equal 

 to four tenths of an inch. Our scale divisions 

 are so large that a Vernier scale is not required 

 for conveying that information. But, assume ~7~ 

 the tenth division between 30 and 31 to be 

 obliterated, still we should be able to discover 

 the overplus beyond 30 inches to be four-tenths 

 by meansof the Vernier scale A, inasmuch as tho 

 number of tenths will be equal to the number of 

 whole parts on the Vernier scale A, above tho 

 first line of coincidence between it and the scale 

 B. Now the line of coincidence in question is at 

 26, counting upwards; from which, to the extre- 

 mity, we have four divisions, which indicate a 

 fraction of four-tenths of an inch over and above 30 inches. 



Correction of the Barometric Column for Capillarity. If 



mercury be poured into a glass vessel, it will not furnish 

 a perfectly level surface, but will be elevated, as in the 

 accompanying diagram (Fig. 13) : it 

 will constitute a meniscus ; and if 

 the line V be dropped perpendicular 

 to the line B, joining the two corners 

 of the meniscus, the line V will con- 

 stitute the versed sine of the menis- 

 cus. The convexity of the meniscoid 

 surface of mercury will vary in pro- 

 portion as the diameter of the tube 

 or other vessel containing the mer- 

 cury varies. Hence, the allowance to 

 be made for diminution of the height 

 of a mercurial co- 

 lumn, owing to ca- 

 pillarity, is deter- 

 mined by two con- 

 siderations the diameter of the tube, 

 and the length of the versed sine of 

 the corresponding meniscus. 



The necessity for such allowance 

 does not apply, as will hereafter be 

 seen, to barometers of every form ; 

 neither is it imperative, so long as 

 the indications of one barometer are 

 to be compared amongst themselves ; 

 but it is indispensable if we would 

 compare the indications of barometers 

 with each other. 



First, let us consider the cases to 

 which the correction for capillarity 

 does not apply- It does not apply to 

 any barometer, the reservoir of which 

 constitutes part of the tube itself. The diagram (Fig. 14) 

 represents a barometer of great tubular diameter. Two 



B.... 



*' 



