356 Proceedings of tlie Royal Society of Victoria. 



as possible, and preferably when the barometer is high, otherwise, 

 possibly, part of the range of the aneroid, and that a most 

 important part, may remain un-tested. While in the chamber, the 

 instrument should be most carefully watched, and its readings 

 noted accurately. It is of vital importance that the index move 

 smoothly and without any sticking or jerking, and that the 

 indications be exactly the same with reducing as with increasing 

 pressure. Friction, and especially irregular friction, of the 

 multiplying mechanism, is a most fatal defect in an aneroid. A 

 gentle tapping or scratching with the thumb nail will set up 

 vibration, and so diminish friction. This is generally used in 

 the field, and some appliance for enabling it to be done in the 

 vacuum chamber is desirable. The readings at every half-inch of 

 mercury having been taken and compared with the barometer, a 

 table, or better still a diagram, of corrections should be made, and 

 used to correct all readings taken in the field. 



In the absence of a vacuum chamber, a fairly good test may be 

 made on a railway that rises rapidly, and has numerous stations 

 at known levels. The railway from Penrith to Katoomba, in 

 New South Wales, which rises over 3000 feet in 30 miles, would 

 answer well. The rack railways, now so common in Switzerland 

 and other mountainous countries, which rise several thousand 

 feet ill a few miles, would be better still. By repeating the test 

 with varying barometric pressures, about 1000 feet more of the 

 range of the instrument may be verified. The same thing may 

 be done for a small part of the range in a tall building such as 

 the Eiffel Tower at Paris, or even the tall business premises 

 existing in Melbourne. The height of the building is best 

 measured by a steel tape line, but by counting the steps of the 

 staircases and measuring the height of a step, a very fair 

 approximation may be obtained. 



In these cases the proper difference of reading is computed 

 from the height l)y means of the appropriate tables, and the 

 actual difference by aneroid being compared therewith, the 

 correction is obtained. It is to be noted that the actual 

 magnitude of the corrections is of no importance. All that is 

 necessary is that they should be accurately known. It is a con- 

 venience to have them all additive. If they are not so, they may 

 often be made so by turning the adjusting screw at the back of 

 the instrument. 



