61 



of the cone in a manner that the observer must learn by trial, the 

 opening at the point may be regulated to any size desired, even 

 while the cone contains mercury. Keep the corie well filled with 

 mercury until all has been added, and do not allow the very last 

 portion to pass through the filter. If the observer has only the 

 supply of pure mercury taken from the barometer, economy must be 

 exercised, but there is no difficulty whatever in being able to filter 

 and utilize the entire quantity of mercury originally in the barome- 

 ter, and this is sufficient. The purity of some of the filtered mercury 

 may be tested as described above. Another indication of the purity 

 of the mercury is the character of the mark left on the paper cone 

 after filtering. To be able to judge by this, observers must filter 

 both pure and impure mercury and compare tlie marks. 



140. The mercury for the cistern, having been filtered at least once, 

 may next be filtered into the cistern, directing the little stream so as 

 to strike against the glass cistern to avoid inclosing small air bubbles 

 near or upon the barometer tube. The open end of the tube should, 

 in the meantime, be completely filled, and the mercury heaped into 

 a little button on the tip end. This button will unite with the mer- 

 cury of the cistern as it rises around the tube, and the chances of 

 inclosing air in the tip end of the tube are thus greatly lessened. 



In general, the cistern should be filled to the brim of the piece i. 

 Before fitting the piece j the leather bag should be pushed out from 

 the inside and every effort used to detach and remove all dust, shreds, 

 little particles of leather, etc. 



141. In securing the clamp rings the screws should be tightened a 

 little at a time, and the precautions cited in paragraph 136 observed 

 to insure a closely fitting and uniformly tight joint. 



When the screws are all tightened, the leather bag should be thrust 

 up into the woden piece j and held there firmly by the finger while 

 the barometer is gradually turned right side up, watching to see if 

 any leaks show themselves at any of the joints. The mercurial 

 column should not be lowered under any circumstances at this time. 

 If a leak occurs, it is probably due to uneven tightening of the joints, 

 and in most cases it is better to loosen the whole joint and shift it 

 a little before tightening again rather than to strain the screws that 

 are already tight, in the hope of making closer contact. 



142. From one to three or more hours after the cleaning operations 

 are completed and the barometer is returned to its support a series of 

 five comparative readings with its companion or standard instrument 

 should be made. 



143. Additional suggestions for Gleaning harovieters with the Tuch 

 cistern. — The parts of the barometers with Tuch cisterns can be 

 removed only by the aid of a special wrench, which will be furnished 

 when any observer is authorized to clean such barometers. 



The special points to be observed are as follows : 



Befor fully screwing up and inverting the barometer, close tightly 

 the air vent d^ Figure 0. Consult also paragraph 17. 



When about to open the cistern, loosen the screw W, Figure 9, one 

 or two turns. Then unscrew the piece q, being very careful at this 

 time not to unscrew the part K. When completely unscrewed, the 

 piece ^, with attached piston c*, may be wholly withdrawn, exposing 

 the mercury. All or nearly all of this may be poured out in the man- 

 ner described in paragraph 127 without clanger of starting the mer- 



