178 Mr. G. J. Stoney on Crookes's Radiometer. 



tally upon it, the vanes move in the direction which indicates 

 an excess of pressure on the blackened sides. Mr. Crookes 

 has measured this excess of pressure ; and its amount, when 

 the light of a standard candle six inches distant was allowed 

 to fall on blackened pith, is stated to have been 0*001772 of a 

 grain upon a surface of two square inches (see an article on 

 " Weighing " a Beam of Light, in i Engineering' of the 18th 

 February, 1876). This is a pressure of somewhat less than 

 the hundredth of a milligramme on each square centimetre. 

 It is the object of the present communication to show that an 

 excess of pressure of about this amount will arise under the 

 operation of known laws. 



3. When a chamber of such a size as to hold the foregoing 

 apparatus is exposed to the prolonged action of a good Spren- 

 gel pump, say for 18 or 24 hours, it is not likely that the 

 outstanding pressure has been reduced lower than about the 

 tenth of a millimetre of mercury ; and as it is not likely that 

 any practicable prolongation of the action of the pump would 

 carry the exhaustion much further, I will assume that the 

 pressure within the chamber is 0*1 of a millim.* This outstand- 

 ing pressure is caused in part by atmospheric air, and in part 

 by the vapours of mercury and of hydrocarbons, to which 

 should perhaps be added vapours of platinum, glass, and 

 other substances. But it will simplify our investigation, and 

 not materially affect our results, to treat it as consisting en- 

 tirely of atmospheric air. 



4. Upon the blackened surface of the vane fell those radia- 

 tions from the candle which are capable of passing through 

 glass. These will have a heating-power not very far from 

 that which Draper's law assigns, according to which the heat 

 that accompanies daylight is proportional to w ly w 2 , w x and 

 iv 2 being the wave-lengths of the longest and shortest undula- 

 tions for which the glass is transparent. These will heat the 

 blackened disk in a considerable degree, but not the transpa- 

 rent glass. I shall assume that the disk is heated one tenth 

 of a degree Centigrade more than the glass. The disk in 

 turn will warm a layer of the air in contact with it. Through- 

 out the thickness of a layer of this kind, if not interfered with, 



* I cannot refrain from observing here how entirely remote such a 

 chamber is from being empty. It follows from what we know of the 

 number of molecules in gases at ordinary pressures, that the number re- 

 maining in this so-called vacuum will be somewhere about a unit-four- 

 teen, i. e. one hundred millions of millions in every cubic millimetre 

 (see Prof. J. Loschmidt, " Zur Grosse der Luftmoleeiile," Academy of 

 Vienna, Oct. 12, 1865; G. Johnstone Stoney on the Internal Motions of 

 Gases, Phil. Mag., Aug. 1868; Sir William Thomson on the Size of 

 Atoms, Nature, March 31, 1870). 



