96 A MANUAL OF TOPOGRAPHIC METHODS. 



attached thermometers should be read. Both barometers should be read by 

 the same observer. A half dozeu observations made at intervals of half an 

 hour will answer as well as a greater number. Such comparisons should, 

 if practicable, be made at the beginning and the end of the season, when- 

 ever a new tube is put into either barometer, or after any repairs to either 

 instrument. 



The discrepancies between the readings of two barometers are due to 

 several causes, among which are differences in setting of the scale of inches, 

 differences in the caliber of the tubes, causing different amounts of capillar- 

 ity and differences in the perfection of the vacuums in the tubes. Differ- 

 ences due to the first two are generally trifling, amounting to but a few 

 thousandths of an inch. If large discrepancies exist, they are usually due 

 to the last cause, and this should be corrected. 



The cistern barometer is a very frail instrument, and although in the 

 mountain form it is protected from accident as thoroughly as possible, still 

 tubes are not infrequently broken while in the field. It is necessary, there- 

 fore, to provide the requisite means for making repairs, such as sealed tubes, 

 distilled mercury, etc. When a tube is broken, the barometer should be 

 opened at once, and the mercury poured out, in order to prevent it from 

 dissolving the screws and other brass work of the instrument. 



The work of filling and replacing a tube is a delicate operation. After 

 taking the barometer to pieces, the new tube should be opened by breaking 

 off the small end, the break being made at a distance from the stricture equal 

 to that upon the old tube. It should be effected by cutting it around with 

 a sharp file, when a little pressure will cause it to break; then the edge of 

 the break should be smoothed with a file. The collar which forms the top 

 of the cistern should then be lashed on to the tube at the stricture. The 

 mercury to be used should be very pure, and to clear it from mechanical 

 impurities, it should be strained through chamois skin immediately before 

 use. It should then be poured into the tube through a paper funnel, and 

 the tube filled to within an inch of the top. Then, covering the open end 

 of the tube with the finger, protected by a piece of kid, invert the tube, 

 letting the bubble of air slowly traverse the tube up and down for the pur- 

 pose of collecting the minute air bubbles which may have remained in the 



