hi) Thermo-electric Currents. 



417 



(lenly in the midst of* a series of experiments, they then pro- 

 duce the leaps which are observed in the curves; in other 

 cases, on the contrary, the alterations are but slowly effected, 

 and they are only detected by causing the elements to pass 

 through the same temperatures. 



I thought it might perhaps be possible to cause these irre- 

 gularities to disappear by avoiding the small quantity of solder 

 which joins the two metals*, and by giving a great section to 

 the worst conducting element. For this purpose I constructed 

 the element represented in fig. 3. A hollow iron tube was 

 curved while red-hot in the direction ABCD; the upper por- 

 tion of this tube was filed away so as to convert the portion 

 BC into an open channel. The extremities A and D were 

 hammered while red-hot in order to cause the interior aper- 

 ture to disappear almost entirely, and in it was incorporated, 

 at a white welding heat, two platinum wires of one millimetre 

 in diameter. These wires are placed in the interior of the hol- 

 low tubes AB and CD; and to isolate them from the iron 

 sides, they were covered with glass tubes ; they terminate in 

 two brass binding-screws, «, b, situated the one alongside the 

 other, and by means of which the element is brought into 

 connexion with the galvanometer. 



This new thermo-electric element was arranged in the ap- 

 paratus in the same manner as in the preceding experiments; 

 the portion which dipped into the bath was kept in a tube full 

 of oil. I will here give the results of some experiments. 



Fifth Series. Sixth Series. 



^' * In the elements employed for the preceding experiments, the two wires 

 of iron and platinum were soldered with silver, but the quantity of silver 

 was inappreciable. 



