440 M. Oersted on [June, 



possible to and beneath this side. By heating one of the soldered 

 parts with a taper, we produced a very sensible effect upon the 

 needle. By heating two soldered places which were not conti- 

 guous, the deviation was considerably increased ; lastly, when 

 the temperature was raised at the three alternating solderings, 

 a still greater effect was produced. We also employed an 

 inverse process, reducing by means of melting ice the temper- 

 ature of one or more of the solderings of the circuit to the freez- 

 ing point. It will be readily conceived that in this process the 

 solderings which are not cooled are to be considered as heated 

 with respect to those that are. This manner of making the 

 experiment admits of ascertaining by different processes the 

 requisite comparisons for discovering the laws of the power 

 investigated. 



Employing the action of ice and that of flame at the same 

 time ; that is to say, by heating the three solderings which were 

 not cooled, we produced a very considerable effect ; the devia- 

 tion amounted to 60 degrees. 



We afterwards continued these experiments with a stronger 

 apparatus, composed of 22 bars of bismuth and 22 of antimony, 

 much thicker than those of the hexagon, and we satisfied our- 

 selves that each element contributes towards the total effect. In 

 order to make some other experiments, we interrupted the circuit 

 in one place, and soldered at the extremities of the separated 

 bars, small brass cups which we filled with mercury, in order to 

 have a ready mode of forming a perfect communication between 

 these two points by means of metallic wires. A copper wire 

 nearly four inches long, and - 04 thick, was nearly sufficient to 

 establish an entire communication ; and two similar wires, one 

 by the side of the other, effected a most perfect communication; 

 a similar wire, about 40 inches in length, also effected a pretty 

 good communication, but a platina wire, 0*01968 of an inch in 

 diameter, and nearly 16 inches in length, occasioned so imperfecta 

 communication that the variation amounted to only one degree. 

 A slip of paper moistened with a saturated solution of soda, com- 

 pletely interrupted the communication. There was no chemical 

 action ; nor did we observe any sensible ignition as might have 

 been expected in an apparatus capable of producing so great an 

 electromagnetic effect. We may also add, that the sum of the 

 effect of all the elements of the complex electromagnetic circuit, 

 is much less than the sum of the isolated effects, which may be 

 produced by employing the same elements to form simple 

 circuits. 



I shall now give a detailed account of the experiments 

 referred to in the above communication, accompanied with some 

 further observations. 



The bars which were employed in the following experiments 

 were parallelopipeds, the transverse section of which was square, 

 each side being nearly 0'6 of an inch in length. 



