400 Prof. A. De la Rive's Researches on the Magnetic 



the liquid column between the poles, and consequently that that 

 should be the distance between the two polar surfaces ; for it will 

 be remembered that, to avoid the influence of the glass disks, 

 each end of the tube containing the liquid entered about 1 cen- 

 tim. within one of the two soft-iron cylinders of the electro- 

 magnet. 



Magnetizing by means of a Bunsen's pile of 60 pairs, I ob- 

 tained, under these conditions, with distilled water, a total rota- 

 tion of 13°; with a pile of from 50 to 40 pairs, the angle was 

 only from 12° to 9°, according to the strength of the pile. 



I will now give the results at which I arrived by operating on 

 some liquids chosen as different as possible from one another. 



Alcohol. 



Five series of experiments, each consisting of several compa- 

 rative observations of rectified alcohol (sp. gr. 0*804) and water, 

 gave the following numbers : — 



Alcohol. Water. Ratio. 



.0 40 



12 10 



0-876 



7 20 



8 20 



0-873 



8 30 



9 40 



0-879 



9 20 



10 40 



0-875 



7 58 



9 5 



0-877 



Mean . . 0*876 



To give an idea of the differences presented by the consecu- 

 tive observations of one and the same series, the numbers above 

 given being the means, I here transcribe in detail those composing 

 the fifth. The first column indicates the azimuth of one of the 

 sensible tints, the second that of the other, and the third the 

 difference between these two angles, which represents the total 

 rotation. The observations upon water alternated with those on 

 alcohol ; but it is easy to perceive that the intensity of magneti- 

 zation did not sensibly vary during the experiment. 



Alcohol. 



16 55 9° 



o / 



7 55 



17 9 5 



7 55 



17 9 



8 



17 9 



8 



17 9 



8 



Mean rotation . 



. 7 58 



