so-called Coercive Force. 519 



this conclusion. This was proved by a simple experiment which 

 I mentioned several years ago, and which showed that two bars, 

 a thick one of iron and a thin one of steel, formed into a cross 

 and suspended horizontally before the pole of a magnet, were 

 directed at a distance by the bar of iron, but near to the mag- 

 net by the bar of steel. Numbers, however, were wanting which 

 I could not obtain from this experiment. The solid bodies 

 examined by me in the memoir referred to were all in the form 

 of powder. It appeared to me therefore very desirable to bring 

 the above question to the test of experiment, operating in a 

 manner which permitted of the examination of iron and steel in 

 a massive form. 



2. For this purpose I had prepared a button of iron (A), and 

 three buttons of steel (B, C, D), as nearly as possible alike in 

 shape and form ; the three latter were made as hard as possible, 

 and afterwards the button B was tempered to blue, the button C 

 to yellow, while the button D retained its full hardness. These 

 three buttons were cut from the same bar of fine English steel, 

 and were rounded hemispherically and polished at tbe ends. 

 Their thickness was 8 millims. ; their length, excluding a small 

 hook for suspension at the other end, was 14 millims. ; then- 

 weight about 6-3 grms. From the same bar of steel, still smaller 

 bars E were cut 53 millims. long, and also hardened and rounded 

 at the ends. The button A was formed from a bar of iron of 

 equal thickness ; and again from the same bar, a smaller one F 

 was taken, equal in size and form to the steel bar E. 



3. Five series of weighings were undertaken in order to deter- 

 mine the weights which were necessary to separate the different 

 buttons from the inducing magnetic pole. For the three first 

 scries of weighings, three of the steel bars E were taken and 

 carefully magnetized in different degrees. These bars were then 

 successively fixed vertically by screws upon a foot ; upon them 

 the different buttons were caused to rest, and the weight which 

 was necessary to effect their separation was determined. In the 

 first series of weighings, in which the button was attracted by 

 the bar which was most feebly magnetized, I made use of a fine 

 Geissler's balance, and the weight was determined by a rider of 

 platinum which was moved along the arm of a lever. In the 

 second series, in which the different buttons were attracted by 

 the bar of medium magnetic strength, the same balance was 

 made use of, but the amount of the attraction was determined 

 by pouring dry sand out of a small orifice upon the scale-pan 

 at the opposite end of the balance-beam. The sand after each 

 determination was weighed with another balance. In the third 

 series of experiments, the glass balance of (i.isslcr had to be 

 changed for a larger chemical balance; the determinations were 



