Art. XXT. — Field Practice with the Aneroid Barometer. 



By professor W. C. KERNOT, M.A., M.C.E. 



[Eead 8th December, 1903]. 



This instrument measures the pressure of the atmosphere by 

 means of an exhausted metallic chamber kept from collapsing by 

 a powerful spring, the motion under varying pressures being 

 magnified by a delicate lever system so as to become easily 

 visible. Tt constitutes a cheap and portable substitute for the 

 more perfect mercurial barometer, and is subject to the same 

 limitations- and also to special defects of its own, which, if not 

 allowed for, render its indications comparatively unreliable. 

 With proper care, however, in choosing, testing and using, it is 

 sufficiently accurate for many purposes connected with explora- 

 tion, surveying and engineering, and vastly more convenient than 

 the cumbrous, fragile and costly mercurial barometer. Extreme 

 differences of opinion have existed amongst experts as to the 

 aneroid, some regarding it as hopelessly unreliable, while others 

 quote examples of astonishingly accurate work done by its means. 

 As is not uncommon in other cases the truth lies between these 

 extremes. The former view is the natural result of neglecting to 

 test the instrument and allow for its errors, which may involve 

 hundreds of feet difference of level, even in good instruments, 

 and which vary enormously in individual cases. For example, a 

 certain aneroid tested at the University gave at one part of its 

 range 1200 feet difference of level, when it ought to have given only 

 1000. Another tested at the same time was almost as perfect as 

 a mercurial at that particular part of its scale, but had an error 

 of 100 feet in 500 at another place. Both these instruments 

 were used most successfully in the field, when these errors were 

 known and allowed for, which without this knowledge would 

 have been almost useless. On the other hand, the extraordi. 

 narily accurate results sometimes quoted for aneroid levelling are 

 no doubt due to good fortune, and the tendency to recollect and 

 quote the successful and forget the unsuccessful instances. As the 

 result of a fairly large experience, it may be laid down that, to 

 determine differences of level up to 100 feet, within 5 feet of the 

 truth, up to 1000 feet within 10 feet, and greater altitudes within 

 1 per cent, of their total amount, is excellent work, while errors 

 of twice the above amount will mark a fair average performance. 



