10 



BAROMETER. 



Fig. 3. 



U^ 



vi 



the sun, or to the currents of the air, which always take place 

 at the joinings of the windows. When the barometer has to be 

 iixed to the wall, as is the case with all the common stationary 

 and wheel barometers, care muot be taken to secure the tube in 

 a position perfectly vertical, regulating it by the plumb-line, first 

 in front, then -at the sides, at least in two vertical planes cutting 

 each other at right angles. When the instru- 

 ment is so constructed as to take its equili- 

 brium itself, as the Fortin barometers and 

 those of J. Green, recently made under the 

 direction of the Smithsonian Institution, it is 

 enough to hang it on a strong hook. These 

 conditions being fulfilled, the rest of the ar- 

 rangement may be varied according to the 

 nature of the localities. For the Fortin and 

 Green barometers, the following arrangement 

 is convenient, and may be almost everywhere 

 adopted. {See Fig. 3.)* 



A small oblong box {a h), some inches 

 longer than the barometer, and a little broader 

 than its cistern, is firmly set against the wall 

 (w w'), near the window, in such a manner as 

 to open in a direction parallel to the panes ; at 

 the summit («) it has a strong hook {h h'), 

 which extends beyond the box about two or 

 three inches, and on which the barometer is 

 suspended. The instrument remains generally 

 in the box, which is closed by a movable cover, 

 and which protects it from external injuries, 

 from dust, and from the direct radiation of 

 warm bodies, or the currents of air from the 

 window, and diminishes the eifect of the too 

 sudden variations of temperature. When it is 

 to be observed, the barometer is taken by the 

 upper end of the tube, and the suspending 

 ring is made to slide towards the end of the 



:> 



* The standard barometer at the Smithsonian Institution is stationary 

 and inclosed within a narrow case, the front and two sides of which open 

 out by means of hinges so as fully to expose the instrument at the time 

 tii the observation. 



