r 444 



XXXII. 



A SIMPLE FOEM OF OPEN-SCALE ISOTHERMAL AIE, 

 BAEOMETER. 



By W. F. BAEEETT, F.E.S. 



[Read February 22. Ordered for Publication May 10. Published May 19, 1910.] 



A LONG-NECKED glass flask filled with air and inverted in a wide-mouthed 

 bottle of water is to be seen in some cottages, where it is used as a weather- 

 glass. If kept at a uniform temperature, the changes in atmospheric pressure 

 will, of course, be indicated by the rise and fall of the water in the neck of 

 the flask. But as this cottagers' weather-glass is ixsually kept on the mantel- 

 piece, the changes of temperature to which it is exposed vitiate its use as a 

 simple form of barometer. If, however, the flask could be rendered impervious 

 to temperature change, it would form an effective and very sensitive 

 barometer. 



It occurred to me some years ago that this isothermal condition might to 

 some extent be attained by using one of Sir James Dewar's liquid air flasks. 

 As is well known, in order to preserve air in the liquid state, Sir J. Dewar 

 devised a glass flask with double walls, the space between being highly 

 exhausted and the walls coated with a reflecting surface of deposited 

 mercury or polished silver. Such a flask was fitted with a thick rubber cork, 

 through which passed a glass quill tube open at both ends, having a liquid 

 index to record changes of volume. Any degree of sensitiveness can easily 

 be attained by altering the ratio between the volume of air in the flask and 

 in the index-tube. In order to protect the latter from sudden temperature 

 changes, the quill tube is encircled with a larger glass tube, closed at its 

 upper end, and having a plug of cotton wool encircling the annular space 

 between the two tubes at the lower end. A scale is now attached, the 

 graduation of which can be easily made by comparing the rise and fall of the 

 index with that of an ordinary barometer when both are enclosed in a 

 receiver and submitted to slight exhaustion by an air-pump. A range of 

 from 6 to 10 inches to I inch of barometric change was found to be con- 

 venient. Bending the index-tube at right angles after it leaves the flask, so 

 that the scale is now horizontal, makes the arrangement more convenient; 

 and to increase the portability of the instrument, the long index-tube can be 

 coiled spirally. 



