Royal Tiistitution, 221 



changes the action on the object, and there are abundance of sub- 

 stances which when surrounded by air are repeUed, and when by- 

 water are attracted, upon the approach of a magnet. When a cer- 

 tain small glass cylinder weighing only QQ grains was submitted on 

 the torsion-balance to the Logeman magnet surrounded by air, at 

 the distance of 0*5 of an inch from the axial line, it required 15° of 

 torsion to overcome the repulsive force and restore the object to its 

 place. When a vessel of water was put into the magnetic field, and 

 the experiment repeated, the cylinder being now in the water was 

 attracted, and 54°'5 of torsion were required to overcome this attrac- 

 tion at the given distance of O'o. If the vessel had contained a 

 fluid exactly equal in diamagnetic power to the cylinder of glass, 

 neither attraction nor repulsion would have been exerted on the 

 latter, and therefore the torsion would have been 0°. Hence the 

 three bodies, air, glass (the especial specimen), and water, liave their 

 relative force measured in relation to each other by the three expe- 

 rimental numbers 15°, 0°, and 54°-5. If other fluids are taken, 

 as oil, aether, &c., and employed as the media surrounding the same 

 glass cylinder, then the degrees of torsion obtained with each of 

 them respectively, shows its place in the magnetic series. It is the 

 principle of the hydrometer or of Archimedes in respect of gravity 

 applied in the case of the magnetic forces. If a different cylinder 

 be employed of another size or substance, or at a different distance, 

 the torsion numbers will be different, and the zero (given by the 

 cylinder) also different; but the media (with an exception to be 

 made hereafter) will have the same relation to each other as in the 

 former case. Therefore to bring all the experimental results into 

 one common relation, a Centigrade scale has been adopted bounded 

 by air and water at common temperatures, or 60° F. For this pur- 

 pose every separate series of results made under exactly the same 

 circumstances included air and water; and then all the results of 

 one series were multiplied by such a number as would convert the 

 difference between air and water into 100°; in this way the three 

 results given above become 21°-6, 0°, and 7S°'4. By such a pro- 

 cess the magnetic intervals between the bodies are obtained on the 

 Centigrade scale, but the true zero is not as yet determined. Either 

 water, or air, or the glass, may be assumed as the zero, the intervals 

 not being in any way dependent upon that point, but the results 

 will then vary in expression thus : — 



all above the zero being paramagnetic, and all below diamagnetic in 

 relation to it. I have adopted a vacuum as the zero in the table of 

 results to be given hereafter. 



In this manner it is evident that, upon principle, any solid, what- 

 ever its size, shape, or quality, may be included in the list, by its 

 subjection to a magnet in air and in water, or in fluids alreadj'* 

 related to these : also that any fluids may be included by the use of 



