Magnetism, Cohesion, Adhesion and Viscosity. 44<9 



in its return to 0°, measured through regular divisions of the 

 compass. On account of the difficulty of starting the needle 

 from 180° when in equilibria, the measurement only began 

 from 175°. 



To simplify the expression of the results, and save the use 

 of decimals, the signs -plus ( + ) and minus ( — ) are used to 

 denote time in excess and paucity, which in the mean of 

 eight observations were too insignificant to be otherwise 

 brought in. 



The mode 1 have described of measuring the insensible 

 magnetic attractions of non-ferruginous bodies, has the rare 

 advantage of the most perfect delicacy and precision in all 

 angles up to 90°, but from thence to the moment when sensible 

 attraction commences, there is considerable vagueness, as far 

 as relative intensities are concerned. Where perfect repulsion 

 occurs it is distinctly indicated by the impossibility of forming 

 a connexion ; but it is likewise evident that in many cases a 

 low degree of repulsion to the magnetic needle is overcome by 

 giving time for magnetic propagation, or by keeping the sub- 

 stance operated upon in contact widi a powerful magnet. 

 Where a body is very light, as cork or charcoal, the failure 

 must be attributed to their want of density, for the part of the 

 needle which comes in contact with the bodies is but a minute 

 point, yet the molecules which compose the cork or charcoal 

 in that spot, must stand at such a distance from one another 

 as to render their magnetism ineffective, and therefore inap- 

 preciable. 



The delicacy of this mode of measurement is sufficiently 

 attested by its having been applied vviUi equal success to a 



Phil. Mag. ^. 3. Vol. 30. No. 203. June 184-7. 2 II 



