58 



tedious, and often the bubble seems to grow smaller and finally dis- 

 appears, being separated into almost imperceptible portions which 

 remain distributed along the walls of the tube. 



Second method. — The following plan is more frequently applied, 

 especially when the quantity of air already in the tube is considerable, 

 is lodged at the top, and must be partially removed at least : 



Empty an inch or two of mercury from the tube. Close the 

 open end tightly with the gloved finger and cause a large bubble of 

 air to glide slowly and regularly along the tube until it unites with 

 all the portions of air it is desired to remove. The large bubble is 

 then as slowly and gradually worked to the open end of the tube 

 again, using every possible precaution to prevent small portions of 

 the bubble from separating and remaining behind. Such a bubble 

 of air may sometimes be successfully passed once into and out of the 

 tube, but even at the best the vacuum in a barometer that has been 

 treated in this manner is very apt to be greatly impaired and can not 

 be restored. The reason of this is that the glass walls of the tube 

 have very strong hygroscopic properties, and while the air bubble is 

 passing along the tube considerable portions of both moisture and air 

 are invisibly retained upon the walls of the tube. While, therefore, 

 a bubble of air may be successfully passed once into and out of the 

 tube, a repetition will be attended with less good effect, as in the 

 meantime the moisture and gases of the bubble will have acted upon 

 the mercury to produce oxidized films that will probably adhere to 

 the walls of the tube, so that when bubbles are again passed there will 

 presently be a marked tendency to cling to the tube and leave small 

 detached bubbles imprisoned against the walls. When, afterwards, 

 the barometer is set up, the walls in the upper portion and near the 

 vacuum, being no longer subjected to the full air pressure as they 

 were while the bubble was passing along the tube, now readily give 

 off both air and moisture, and in many cases numerous little bubbles 

 form against the walls even below the top of the column and probably 

 later work their way into the vacuum. 



133. The removal of air from a barometer tube, therefore, can not 

 be perfectly effected in any such way, and should not be undertaken 

 unless the defect is a very serious one. If the comparative readings 

 taken before cleaning a barometer do not show serious errors, any air 

 the tube may be thought to contain had best be allowed to remain. 



134. One of the most difficult and delicate parts of the process of 

 cleaning is that about the wooden piece G and ivory point. The deep 

 and narrow annular space between the glass tube and the boxwood is 

 generally covered with oxide of mercury, which should be thoroughly 

 removed by repeated wiping with clean cloths applied upon the 

 ends of slender sticks or by similar means. Tufts of raw cotton 

 will adhere firmly to, and are readily wrapped about rough sticks 

 and may serve with advantage in wiping out the narrow spaces. 

 Sometimes, however, the space is so small that it can not be properly 

 cleaned. Care must be observed not only here but in subsequent 

 operations to blow away or otherwise remove every vestige of lint, 

 dust, shreds of cotton, etc., as, if allowed to remain about the parts 

 of the cistern, they will quickly find their way to the surface of the 

 mercury, upon which they will float about to the detriment of accu- 

 rate adjustments. 



