70 ■ Royal Society. 



be shown by the gauge as soon as the magnet was rendered active; 

 but whatever gas was employed, or whatever power of magnet used, 

 not the slightest change was produced. 



Thinking it possible that there might be expansion in one di- 

 rection and contraction in another, the gases were then examined as 

 to the production of any currents in them, but no traces of such 

 appeared. 



From these results, the conclusion was arrived at by the author, 

 that the motions of magnetic and diamagnetic bodies in each other 

 do not appear to resemble those of attraction or repulsion of the ordi- 

 nary kind, but to be of a differential action, dependent perhaps upon 

 the manner in which the lines of magnetic force were affected in 

 passing from one to the otiier during their course from pole to pole ; 

 the differential action being in ordinary cases between the body 

 experimented with and the medium surrounding it and the poles. 

 A method of showing this action with the gases is described, in 

 wiiich delicate soap-bubbles are made to contain a given gas, and 

 these, when held in the magnetic field, approach, or are driven further 

 off, according as they contain substances, magnetic or diamagnetic, 

 in relation to air. Oxygen passes inwards or tends towards the mag- 

 netic axis, confirming the results formerly described by the author in 

 Ills account of his investigations of flame and gases. 



Perceiving that if two like bubbles were set on opposite sides 

 of a magnetic core or keeper cut into the shape of an hour- 

 glass, they would compensate each otiier, both for their own dia- 

 magnetic matter and for the air which they would displace; and 

 tirdt only the contents of the bulbs would be virtually in a differ- 

 ential relation to each other, the author passed from bubbles of 

 soapy water to others of glass ; and then constructed a differential 

 torsion balance to which these could be attached, of the following 

 nature : — A iiorizontal lever was suspended by cocoon silk, and at 

 right angles, at the end of one arm, was attached a horizontal cross- 

 bar, on which, at about 1^ inch apart, and equidistant from the 

 horizontal lever, were suspended the glass bubbles ; and then the 

 whole being adjusted so that one bubble should be on one side of 

 the iron core and the other on the other side, any difference in their 

 tendency toset inwards or outwards from the axial line caused them to 

 take up their places of rest at different distances from the magnetic 

 axis; and the power necessary to bring them to an equidistant position 

 became a measure of their relative magnetic or diamagnetic force. 



In the first place, different gases were tried against each other, and 

 when oxygen was one of them it v^ent inwards, driving every other 

 outwards. The other gases, when compared together, gave nearly 

 equal results, and require a more delicate and finished balance to 

 measure and determine the amount of their respective forces. 



The author now conceived that he had attained to the long-sought 

 power of examining gaseous bodies in relation to the effects of heat 

 and the c ftects of expansion separately ; and proceeded to an investi- 

 gation of the latter point. For this purpose he prepared glass bubbles 

 containing a full atmosj)here, or half an atmosphere, or any other 

 jjroportion of a given gas ; having thus the power of diluting it 

 without the addition of any other body. The effect was most stri- 



