Mr. J. D. Fry on a New Micromanometer. 495 



its construction, and its calibration are discussed below ; but 

 it mav be mentioned at this point that in its most delicate 

 form the new manometer possesses a sensitiveness at least 

 ten times that o£ any micromanometer previously described. 



The displacement o£ a stretched membrane is dependent 

 on (1) its tension ; (2) its stiffness. In order that the dis- 

 placements may be proportional to the pressures applied, it 

 is important that of these two factors the tension should 

 predominate. To achieve this it is necessary to use a very 

 thin membrane and to arrange that the tension to which it 

 is subjected is considerable. 



If the stiffness may be neglected and the displacement be not 

 too great, the tension will remain constant : the displacement 



TOT 



is then equal to ^-, where p is the pressure difference 



between the two sides, r the radius of the membrane, and T 

 the tension in the membrane. 



The selection of the membrane to be used will depend on 

 the order of magnitude of the pressure differences which the 

 manometer is required to measure. The following substances 

 have been tried : — 



Thin sheet-rubber. — This was the first substance to be used. 

 It provides a membrane of extreme sensitiveness, but it is 

 not possible to put it under sufficient tension to eliminate 

 entirely a drift of zero, and it is for this reason unsatis- 

 factory. 



Goldbeater s skin. — This has a thickness of approximately 

 0'0025 cm., but cannot be obtained in sufficiently large sheets 

 to be of much service. 



Cyclostyle paper. — Being both thin and impervious to air, 

 cvclostyle paper is a satisfactory membrane when the highest 

 .sensitiveness is not required. 



Thin copper. — Copper sheets cannot be obtained much 

 thinner than 0"004 cm., but membranes of this thickness, 

 or even greater, may be employed with advantage for the 

 measurement of comparatively large pressures. 



" Badiselie." — For the indication of the smallest pressures 

 this is so far the most suitable membrane which has come to 

 the author's notice. It is a thin animal membrane, and has 

 been supplied by Messrs. Ferris k Co. of Bristol. It may 

 be obtained in sheets about 25 cm. square and is considerably 

 thinner than goldbeater's skin, some sheets having an average 

 thickness of about 0*001 cm. By coating this membrane 

 with a thin petrol-resin varnish it is rendered impervious to 

 air. 



