PHYSICS. 99 



rated plate, and below which is a three-way cock with a 

 lateral tubulure, through which the gas is introduced. Two 

 marks on the tube are the points between which the effusion 

 time is noted, the gas being forced out through the plate by 

 a water or mercury column in the open leg of the tube. The 

 results obtained with the instrument were accurate for coal- 

 gas and oxygen, but varied widely for hydrogen. 



Kraevitsch has proposed an improvement in the construc- 

 tion of the barometer which increases indefinitely the sensi- 

 bility of this instrument. To the shorter leg of a siphon 

 barometer is attached a long horizontal capillary tube ter- 

 minating in an open cylinder, the space above the mercury 

 and the capillary tube being filled with water free of air. 

 Obviously, if the barometer rises or falls, a quantity of wa- 

 ter is displaced by the mercury equal to the volume repre- 

 senting the change in height. If now a bubble of air be in- 

 troduced into the capillary tube, it will be displaced by an 

 amount equal to the change in the barometric height, multi- 

 plied by the ratio of the two sections in Kraevitsch's in- 

 strument by 140; thus rendering it extraordinarily delicate. 

 The bubble when observed by a microscope of low power 

 is rarely in repose. 



An extended posthumous paper upon the constants of 

 aneroid barometers and upon those aneroids which have 

 scales attached for measuring heights, by Professor Jelinek, 

 of Vienna, has appeared. It contains a complete resume of 

 previous results obtained by various observers. 



ACOUSTICS. 



Lord Rayleigh has experimented to ascertain the maxi- 

 mum limit of the amplitude of sound-waves, using for this 

 purpose a whistle mounted on a Wolfe's bottle, furnished 

 with a manometer. It was found that the most suitable 

 pressure was 94- centimeters of water, and that under these 

 conditions the sound could be distinctly heard at 820 me- 

 ters' distance. The amount of air passing through the 

 whistle was found to be 196 cubic centimeters per second. 

 From these data the required amplitude may be readily 

 calculated. The result shows that the amplitude of vi- 

 bration of the aerial particles was less than the ten-mill- 



