57 



titles of mercury will be spilled into or remain in various little cracks 

 and crevices while the cistern is being emptied. These, by all means, 

 should be thoroughly dislodged, especially from about the metal 

 parts. With the glass tube removed, the sheath should be tapped and 

 shaken smartly to remove all small globules of mercury. It may then 

 be wiped and cleaned thoroughly with cloths or chamois skin. In 

 case the scale is somewhat dull and tarnished it may be brightened 

 by suitable polishing, but this is a delicate operation and should be 

 avoided rather than otherwise. The danger lies in shifting the posi- 

 tion of the scale, and if polishing is absolutely necessary, it should, 

 therefore, be done with very great care. 



The upper portion of the glass tube should also be cleansed on the 

 outside with the aid of a damp cloth if necessary. 



129. Air in barometer tubes. — How air can gain entrance to the 

 vacuum of a barometer otherwise in good condition, which is sup- 

 posed to have been hanging quietly and undisturbed upon its sup- 

 ports, is a matter that is very difficult both to imagine and to explain. 

 No case of this sort has ever occurred among the hundreds of 

 barometers handled at this office, so that when such a defect is dis- 

 covered in an instrument in use at a station the observer in investi- 

 gating the cause and reporting the matter should make sure that 

 the barometer has not been tampered with or roughly handled by 

 unauthorized persons, as, if uninjured in other respects, misusage is 

 the most probable explanation of the defect. 



130. If an appreciable quantity of air is in the tube at the time of 

 cleaning, it can be seen more or less conspicuously in the shape of a 

 small bubble or bubbles adhering closely to the walls of the tube. 

 If these bubbles appear no . larger than good-sized pinheads, and 

 especially if they are not more than halfway up the tube, then it is 

 certain that the condition of the vacuum is more likely to be greatly 

 impaired than improved by attempts to remove them. 



131. Sometimes the barometers that observers may be calleci upon 

 by private parties or friends to inspect or repair, seem to have numer- 

 ous rather flat-shaped air bubbles firmly lodged against the sides of 

 the tube. Generally these are not air bubbles at all, but are parti- 

 cles of moisture, the presence of which is due to carelessness in the 

 original preparation and filling of the tube. The edges of an air 

 bubble are sharp and the mercury generally remains bright and 

 makes well-defined contact at a steep angle with the glass. If some 

 moisture is present, either alone or with the air, the edges are less 

 clearly defined, the mercury is oxidized, and the angle of contact is 

 less steep, the bubble itself being very flat. 



It is impossible, without entire cleansing, drying, and refilling, to 

 do anytjiing with a barometer that contains moisture. 



132. If a bubble or so of air is present in a tube, the plan that 

 should first be tried to remove it is as follows : 



First method. — Incline the tube 45° or thereabouts, with the open 

 end up, and tap it gently in the vicinity of the bubble, revolving the 

 tube a little at the same time so as to encourage the bubble to creep 

 along the inclined surface of the glass. If the inclination is too 

 great the bubble will be greatly compressed by the weight of mercury 

 above it ; if too small the bubble will not tend to move. 



If the treatment is successful and the bubble is removed, the result 

 will probably be beneficial ; but at best the operation is generally very 



