[ 415 ] 



\ 



LIII. The Measurement of Small Gaseous Pressures. 

 By Charles F. Brush*. 



PRIOR to the invention of the McLeod vacuum gauge, 

 the measurement of even moderately small gaseous 

 pressures was difficult, and subject to large errors. The 

 introduction of the McLeod gauge, however, early in the 

 seventies, seemed to solve the problem. In its ordinary form, 

 and for most purposes, this beautiful instrument admirably 

 serves the purpose for which it is designed. But when very 

 accurate measurements of pressures as small as a few millionths 

 only, of atmospheric pressure are desired, its performance is 

 extremely unsatisfactory and vexatious. As is well known, 

 the chief cause of the difficulty is the unequal and variable 

 capillary depression of the two small columns of mercury, 

 whose difference in height indirectly serves as the measure of 

 pressure. Accurate measurement of this capricious difference 

 obviously avails nothing. 



Three or four years ago, I was engaged in an investigation 

 requiring frequent and simultaneous measurements of slight, 

 but different pressures, in two large glass globes connected 

 by a capillary tube. For this purpose I constructed and 

 carefully callibrated two large McLeod gauges. The internal 

 diameter of the mercury tubes was about 3 millimetres, 

 and they were made from contiguous parts of the same glass 

 tube selected for uniformity of bore. These gauges were 

 often compared by measuring the same vacuum with both ; 

 but they rarely gave concordant results. Indeed it was not 

 uncommon at high exhaustions, for one or the other of them 

 to indicate a negative vacuum ; that is to say, less than no 

 pressure at all. The case of these two gauges is cited because 

 of the opportunity they afforded for comparison. In prior 

 work I had, like most experimentalists, used but one gauge ; 

 and while always suspicious of its indications, had no means 

 of knowing how large its errors might be. 



The phenomenon which I next desired to investigate, was 

 the spontaneous evolution of gas from glass, and other surfaces 

 in high vacua. For this purpose an accurate and entirely 

 reliable means for measuring very small pressures was neces- 

 sary, because I could not afford to wait months or years for 

 the evolution of sufficient gas to be detected with certainty 

 by the old gauges. To meet these requirements I designed, 

 constructed, and learned how to use the modified form of 



* Communicated by the Author, having been read before the American 

 Association for the Advancement of Science, August 12, 1897. 



