INTENSITY OF MAGNETIC AND DIAMAGNETIC FORCES. 745 



or depressed below the line of the apices of the poles, they deci- 

 dedly assumed an equatorial position, as a strongly diamagnetic 

 uncrystalline body of the same form would have done, I shall 

 again recur to this subject on some future occasion, for the 

 present I must leave it/ It is however so far certain, that the 

 ferridcyanide derived from different sources is magnetic, 



47. The magnetism of the ferridcyanide, in opposition to the 

 diamagnetism of the ferrocyanide, is the more remarkable, as the 

 latter (FeCy + 2KCy) is a compound of the protocyanide of iron, 

 and the former (Fe^Cy^ + SKCy) of the percyanide of iron with 

 cyanide of potassium ; whilst the percyanide is a form of com- 

 bination in which the amount of iron compared with the cyano- 

 gen diminishes in a greater proportion than in the protocyanuret. 

 In a certain sense, the different reaction of the proto- and per- 

 chloride of iron, in which case (in solution) the latter is certainly 

 not magnetic, but less diamac/netic than the former, forms an 

 analogy to this. 



The magnetism of the ferridcyanide appears to be too great 

 to suppose, as occurred to me for a moment as probable, that 

 it might be ascribed to an admixture of protochloride of iron, the 

 amount of which would then be too great. 



48. Finally, if we glance at the last column of the table in the 

 4 1st paragraph, which,ybr equal weights, gives the diamagnetism 

 of the different substances, it is evident that this diamagnetism 

 in all the substances enumerated, which are not mixtures in in- 

 definite proportions, may be expressed within the limits of error 

 of observation by perfectly simple numerical relations. The 

 greatest deviation, one-nineteenth, occurs in sublimed sulphur 



I and chloride of sodium ; but even in these cases the probable 

 error is greatest, on account of the want of uniformity in pressing 

 in the powdered substance. The simple numerical relations 

 alluded to are : — 



Phosphorus, water, sulphuret of carbon 1, 



and hydrochloric acid J 



Sulphuric a?ther and oil of turpentine . | 

 Sublimed sulphur and chloride of sodium f 



Nitric acid *- 



Oxide of bismuth and sulphuric acid . ^ 



Mercury j 



Are these relations accidental, or will they be generally con- 

 i firmed? We must wait and see whether the latter occurs. 



