1878.] of Iron and Steel caused by Magnetism. 



113 



A 



M.C. not flowing. 

 1st minute 40 



B 



M.C. flowing. 

 1st minute 34 



C 



M.C. not flowing. 

 1st minute 31 



D 



M.C. flowing. 

 1st minute 28 



2nd minute 39 

 3rd „ 38 

 4th „ 34 

 5 th 36 



tj 111 35 t-r\J 



2nd minute 34 

 3rd „ 33 

 4th „ 31 

 5th „ 32 



2nd minute 33 

 3rd „ 31 

 4th „ 31 

 5th „ 30 



2nd minute 29 

 3rd „ 27 

 4th „ 29 

 5th „ 28 



No. of divisions in 

 last 4 minutes 

 = 147. 



No. of divisions in 

 last 4 minutes 

 = 130. 



No. of divisions in 

 last 4 minutes 

 = 125. 



No. of divisions in 

 last 4 minutes 

 = 113. 



The light was then near the other end of the scale, so the adjusting 

 magnet was employed to bring the light back again. The set of 

 experiments here given, the first of many, is rather an unf avourable one, 

 most of them giving results much closer together than those here given, 

 and, in several instances, the flow of heat would seem perfectly steady 

 for upwards of fifteen minutes. 



The observations of the first minute in each case were not taken, in 

 order to avoid error from very slight deviations of the galvanometer 

 needles caused by the action of the electro-magnet, which however was 

 placed at such a distance from the galvanometer, and in such a position 

 as not, in most cases, to produce any such deflection, and to give time 

 for the magnetism to produce its effect on the bar. It has been de- 

 termined, that in experiments made in this manner, 1 minute seems 

 quite sufficient for the above-mentioned purpose. 



Taking A and C together, we obtain as a mean the number of 

 divisions passed over in 4 minutes with the M. C. not flowing 136 as 

 against 130 from B, with the current flowing ; again, from B and D, 

 with the current flowing, we obtained 121*5 as against 125, with the 

 current not flowing, so that, in both cases, there is a less mean flow 

 with the M. C. flowing than when it is not flowing. The mean of this 

 particular set of experiments would give a decrease of flow, when the 

 bar is magnetised longitudinally, of about 3*6 per cent, of the whole 

 for a magnetising current, causing a deflection of the needle of the 

 galvanometer of 18* 6°. The mean of all the observations for the 

 specimen, and for this current, gave a decrease of flow amounting to 

 3 3 per cent, of the whole. 



The electro-magnet was now turned through 90°, so as to magnetise 

 the iron transversely, and a similar set of experiments were made. 



Here, again, the result was most conclusive, and the mean of several 

 observations showed an increase of flow when the bar was magnetised 

 transversely, amounting to about 3'2 per cent, of the whole for the 

 same current strength. 



Thus, the decrease of conductivity in one case seems roughly to be 



VOL. XXVII. i 



