12 



lid of which closes with a spring clasp, and not only secures the 

 barometer from accidental damage but from undue exposure to 

 atmospheric influences as well. 



32. How haro'nietm' tubes may he -piled. — Processes that may be 

 followed in filling barometer tubes for high-grade instruments are 

 so rarely described and so little known that a short description of 

 some methods frequently employed at the Weather Bureau with 

 highly satisfactory results will doubtless be of interest to a number 

 of readers. 



The object of any filling process is simply to introduce pure mer- 

 cury and totally exclude all air, moisture, or other foreign matter 

 especially of a gaseous nature, that may possibly later occupy the 

 barometric vacuum and cause errors by the pressure it exerts upon 

 the top of the mercurial column. 



33. Funnel method. — It may seem that the desired result could be 

 obtained by carefully introducing clean mercury through a long 

 slender-stemmed funnel reaching quite to the closed end of the tube, 

 (See fig. 13.) A suitable funnel may easily be made by drawing 

 down the end of a short piece of rather wide glass tubing. Such a 

 method is sometimes used and will, indeed, give approximate results, 

 but it will be found upon investigation that while the mercury seems 

 to drive out all the air, yet a good deal will still be found in the 

 vacuum. Originally, this air mixed with water vapor is strongly 

 adherent to the walls of the glass tube by reason of a peculiar prop- 

 erty this character of glass is found to have. When the barometric 

 vacuum is formed, some of the gaseous matter thus attached to the 

 tube is liberated and by its pressure depresses the mercurial column 

 several hundredths of an inch, as has been shown by careful experi- 

 ments. 



In all cases it is of great importance that the inside walls of barom- 

 eter tubes be perfectly clean. New tubes are thoroughly cleaned 

 with whiting or other suitable material while open at both ends, and 

 while still warm and dry the top end is closed and the cistern end 

 tapered and finally fused shut. 



Small tubes (one-fourth inch and less) that have become soiled by 

 use, exposure, etc., can not be easily cleaned properly, and such are 

 never used a second time in the Weather Bureau work. The methods 

 given in paragraph 35 for cleaning larger tubes may, however, be 

 used even with these. 



One very simple and excellent method of driving off nearly all the 

 air and moisture condensed on the glass walls is given in the next 

 paragraph. 



34. The boiling method. — This is a simple method commonly 

 employed with all small tubes, say, one-fourth inch diameter, more 

 or less; such, for example, as is required in the several types of 

 barometers that have been heretofore described. Much larger tubes 

 are frequently boiled, but these when full of hot mercury are diffi- 

 cult to handle, a strong heat is required, and the danger of serious 

 accidents is considerable. 



It is well at first to warm more or less the whole tube, and the cup 

 of clean filtered mercury ^ from which the supply is drawn should 

 also be gently warmed. 



1 For notes and remarks on testing and filtering mercury, see pars. 137 and 138. 



