482 Scientific Proceedings, Royal Dublin Society. 



by the gas present. If, however, means were employed for keeping 

 up the oscillations, as, for instance, by attaching a small piece of 

 iron to the ball, and attracting it when in the middle of its swing 

 by means of a magnet, or by moving the point of suspension, or by 

 other means, the period of oscillation could be determined to any 

 desired degree of accuracy. The complete theory of the experi- 

 ment is, however, very complicated. The highest mathematics 

 have been expended, on calculating the efEect of the gas present on 

 the period of oscillation of a pendulum ; and although the theory 

 of a sphere performing oscillations round one of its diameters is 

 enormously simpler, yet the way in which its period of oscillations 

 depends on the viscosity of the gas is rather complicated, and this 

 would make it a rather complicated operation to calculate accurately 

 the density of the gas from observations with this instrument. It 

 might, however, be used for rough observations, and as the same 

 series of observations by giving the logarithmic decrement of the 

 amplitudes of oscillations would give a measure of the viscosity of 

 the gas, it would be an interesting instrument to make observa- 

 tions with on a considerable series of gases and vapours. 



The second instrument I constructed works upon the same 

 principle as an old philosojahical instrument called a " Baroscope", 

 It was constructed for the purpose of demonstrating the amount by 

 which bodies are buoyed up by the air in which they are immersed. 

 A piece of cork or a glass bulb is balanced in a rough balance by 

 a lead weight. If this be placed under the receiver of an air-pump, 

 the lead weight no longer balanced the cork or bulb, as it is no 

 longer assisted by the presence of dense air. A delicate apparatus 

 of this kind is evidently capable of exactly the same accuracy as 

 the ordinary method of weighing the gas in the bulb, as in Dumas' 

 method, for it is only weighing the gas displaced by the bulb in- 

 stead of that in it. As, however, most delicate balances are large, 

 I thought it might be well to construct an apparatus of this kind 

 on a very small scale, so as to be able to deal with comparatively 

 small volumes of a gas. I have, therefore, constructed one as 

 follows : — I blew a thin glass bulb of about 1 cm. diameter on a 

 fine quill tube which I drew out to a thinner stem about 2 cm. 

 long. To the end of this I attached a mirror such as is used for 

 galvanometers with its plane at right angles to the stem. At a 

 point in the stem close to the bulb I attached two very small and 



