THE MANOMETERS. 31 



If it is desired to move the thread to the right, the rubber tube (e), 

 to the left of the ball (d), is compressed between the thumb and fore- 

 finger of the left hand until the meniscus has taken the right position 

 under the miscroscope, when, without releasing the tube, the rubber over 

 the ball (d) is pinched between the thumb and forefinger of the right 

 hand until a passage for air is opened. The portion of the rubber tube 

 which is held in the left hand may then be released, since any difference 

 in atmospheric pressure at the two ends of the thread is quickly equal- 

 ized through the passage which has been opened over the ball (d), and 

 without disturbing the thread. If the thread is to be moved to the 

 left, the rubber tube to the right of d is compressed between the fingers 

 of the right hand, and the passage for air over the ball is made with the 

 left hand. After a little experience, the exact adjustment of the cali- 

 brating thread^becomes easy and nearly automatic. 



Fig. 17. — Second arrangement for calibrating manometers. 



(A) Tube to be calibrated (bore from 0.45 to 0.65 millimeter); (B) rubber tube; (a, a) bulbs 

 blown on each end of tube to prevent loss of calibrating thread; (b, b) lines of reference 

 etched on tube; (c) calibrating thread; (d) steel ball for setting calibrating thread; (e) rubber 

 tubing; (J, f) glass tubes; (g, g) rubber tubes. 



The calibration is commenced somewhat below the lower scratch — 

 the etched line to the left — and consists, as when the tube is calibrated 

 in the vertical position, in setting the thread exactly end to end and 

 determining its length until the thread has passed the upper scratch. 

 It is then run out of the tube and weighed. Afterwards the whole 

 of the calibrated portion of the tube is filled with mercury, which is 

 also run out and weighed. 



From the length and weight of the long thread, the mean diameter 

 of the bore is calculated; and from the observations on the length of 

 the short thread in the different parts of the tube, a mean calibration 

 unit is derived, and a curve of corrections constructed, exactly as in 

 the calibration of a eudiometer. Finally, a mean value for the double 

 meniscus is obtained from the length and weight relations of the long 



