518 



Mr. (Jr. Gore on 



and drained as completely as possible, heated to redness, coiled, and 

 placed in position ; the galvanometer connected ; the vessel charged 

 with the clear and cold solution by means of the funnel, until 

 the electrodes were quite submerged; and the deflection of the 

 galvanometer needles was also noted. The cistern was then charged 

 with water ; and as soon as the needles of the galvanometer settled 

 at or near zero, a Bunsen's burner (previously placed beneath the 

 projecting chamber K) provided with a regulating tap, was lighted, 

 the water heated to boiling, and boiled until the temperature of the 

 solution ceased to rise, the deflections of the needles during the 

 process of heating being noted. The hot water was then ran out, 

 the apparatus allowed to cool, then discharged of its solution, and 

 thoroughly washed by pouring abundance of cold water through 

 each limb. It usually required about three or four minutes to 

 raise the water in the cistern to 212° F.; the liquid in the cup C 

 usually acquiring a temperature of nearly 150° F. by that time; and 

 it required the water in the cistern to be kept boiling about six or 

 eight minutes longer in order to raise the temperature of the inner 

 liquid to 200° F., and that was nearly the highest temperature it would 

 acquire. 



This apparatus acted very satisfactorily, and enabled me to more 

 accurately examine the influence of heat in such actions. The results 

 obtained with this form of thermo-electric liquid examiner, confirmed 

 in nearly every case, where the currents were feeble, those obtained 

 with the other one, and confirmed also so far as regards the chief 

 results (viz., the existence and direction of the current) those obtained 

 with strongly acid or alkaline liquids, but less frequently confirmed 

 with such solutions the relative magnitudes of the currents in diffe- 

 rent cases ; the large cylinder apparatus may therefore be satisfactorily 

 employed with neutral solutions, but is not suitable for purposes of 

 accurate measurement of the currents, especially of those obtained 

 with strongly acid or alkaline liquids. 



Although the experiments made with the cylinder apparatus were 

 liable to interferences caused by the washers, as nearly all of them 

 gave substantially reliable results, it was desirable to record them in 

 the following pages for the purpose of comparison. A sufficient 

 number of liquids was examined by means of each apparatus so as to 

 include apparently exceptional cases as well as conspicuous and extreme 

 instances of the phenomenon; and the following is an abbreviated 

 statement of the results obtained, with occasional remarks upon them. 



Behaviour of Different Liquids and Metals. 

 A. — In the Cylinder Apparatus. 

 In all the experiments with this apparatus, the water in contact 

 with the upper plate was raised to a temperature of about 210* F., 



