82 Dr. Tyndall on the Progress of the Physical Sciences : 



jealously test every result of liis predecessors. This is the state 

 of things which the investigation of M. Magnus is intended to 

 remedy, and his memoir on the subject furnishes internal evi- 

 dence of the precision with which the inquiry has been con- 

 ducted. The investigation is far from exhausting the subject, 

 but it lets us know precisely where we are ; new and striking- 

 facts have been added, errors have been corrected, anomalies 

 accounted for, and the first great step made towards the reduc- 

 tion to law of those inexplicable phsenoinena which have hitherto 

 perplexed philosophers. 



The wire usually applied in the construction of galvanometers 

 often presents a difficulty in inquiries like the present. That 

 purchased at the merchants is so magnetic as greatly to interfere 

 with the parity of the experiments. To obviate this defect, 

 some precipitated copper was obtained from a galvano-plastic 

 manufactory ; but the metal, after having been cast into cylin- 

 drical moulds, was found so magnetic as to necessitate its re- 

 jection. The pure metal was finally obtained in the following 

 manner : an excess of ammonia was added to a solution of sul- 

 phate of copper, the precipitated oxide being thus redissolved, 

 and the iron mixed with the salt separated ; the solution was 

 filtered, evaporated to dryness, and the ammonia expelled ; the 

 sulphate thus procured was redissolved in water and precipitated 

 by the voltaic current. This metal, however, was exceedingly 

 brittle, and required to be melted eight times in succession be- 

 fore it could be drawn into wire ; when drawn, however, it was 

 found to answer its purpose perfectly. 



In the following pages we shall often have occasion to speak 

 of the direction of the current, and it is therefore prudent to 

 define clearly in the first instance what is meant by this expres- 

 sion. If a strip of copper and a strip of zinc be immersed in a 

 conducting fluid, and the exposed ends be united by a copper 

 wire, the current is said to proceed from the copper through the 

 uniting wire to zinc, and hence from the zinc through the fluid 

 to copper. Suppose a bit of antimony to supersede the strip of 

 copper, and a bit of bismuth in the place of the zinc, and doing 

 away with the fluid, let the free ends of both be brought into 

 contact and the place of contact heated ; the consequent thermo- 

 current will act upon a magnetic needle exactly as that developed 

 by the zinc and copper pair. The current therefore passes from 

 antimony through the wire to bismuth (from A to B), but from 

 bismuth to antimony (against the alphabet) across the place of 

 junction. Whenever it is stated in this Report that the current 

 passes from one metal to another, the words " across the place 

 of junction " are always implied. 



It was soon ascertained that a difference in point of hardness 



