668 CONSTANTS OF MAGNETISATION. 



least magnetisation is less or greater than 45; that is, according as 



or 



1229. The direction which a crystal takes in a uniform field 

 only depends on the difference of the coefficients ; we may then 

 add a constant value to each without modifying the phenomena. 

 From this it follows that the conditions of equilibrium are not 

 changed when the body is immersed in any given medium, mag- 

 netic or diamagnetic.* 



If there were bodies (397) whose coefficients /, /', and /", were 

 not all of the same sign, the body being magnetic in certain direc- 

 tions and diamagnetic in others, the properties might still be repre- 

 sented by an ellipsoid of induction which would be obtained by 

 adding a constant to all the coefficients. 



This indifference of the medium does not exist for a non- 

 uniform field, for an element of volume is attracted towards points 

 of maximum or minimum force, according as the coefficient of 

 apparent magnetisation is positive or negative. Faraday observed, 

 for instance, that a crystal of ferrid-cyanide of potassium, varnished 

 on the surface so as to prevent its being dissolved, is constrained to 

 move in the direction of increasing forces when it is immersed in 

 water, and, on the contrary, moves towards decreasing forces when it 

 is immersed in a concentrated solution of ferrous sulphate. But in 

 a liquid formed of 15 volumes of the concentrated solution, to which 

 6 volumes of water have been added, the crystal is magnetic or 

 diamagnetic according as its axis of symmetry is parallel or perpen- 

 dicular to the lines of force that is to say, that in the first case it 

 moves towards increasing, and in the second towards decreasing, 

 forces. 



Tyndall and Knoblauchf have observed that we can obtain 

 analogous figures with bodies in powder compressed in a definite 

 direction, or of systems of superposed layers. The direction of 

 greatest density always tends to set parallel to the lines of force, 

 and perpendicularly to those lines for diamagnetic bodies. This 

 experiment is made by a paste of gum, in which is incorporated 

 powdered bismuth, or grains of carbonate of iron. 



* FARADAY. Experimental Researches, Series, xxn. and xxx. 1855. 

 t TYNDALL and KNOBLAUCH. Phil. Mag., Vol. xxxvi., p. 178. 1850. 



