126 Proceedings of the Royal Society of Victoria. 



example of this I ever obtained, but I could never again get one 

 nearly so good. The time occupied in taking the observations 

 from which it is drawn was about two hours. When the flame 

 was shifted forward a centimetre, the needle crept slowly to the 

 next position, there not being a single oscillation the whole time. 

 For a long time I had been unable to obtain more than -001 volt 

 from German silver, but on one occasion I found two different 

 parts of a wire, one of which gave '0014, the other -0016, giving 

 the same values after several heatings. I then heated them 

 simultaneously and got -0031, which afterwards increased to 

 •0042, and after being heated and cooled several times increased 

 to "0047, the wire then fused. On another occasion, after heating 

 a wire for some time, with little effect, I left the flame alone for 

 a considerable time, and the electromotive force rose steadily to 

 •0045, and unsteadily to •0052, and then fell as suddenly as if 

 the wire had parted, but it had not, for on the oscillations of the 

 needle dying out there was still a small deflection. 



To examine lead I first of all dipped part of a lead wire into 

 hot oil, but could get no effect. I afterwards melted some lead in 

 the bowl of a pipe and heated the stem so as to make it run along 

 and fill it up. The lead could thus be heated far above its melting 

 point without running away and breaking circuit. On heating 

 this tube the effect was apparent at once, and on irregular 

 heating soon became considerable, '001 volt. The tube was then 

 heated systematically from end to end, and it was found that 

 there was about an inch towards one end which always gave 

 great results when heated, while at all other parts the effect was 

 very small. After several heatings •OlS volt was observed at 

 this critical point. Though the movements of the galvanometer 

 needle were very slow and generally dead beat, yet, about this 

 part, there were some peculiar effects. On one occasion, on 

 applying the flame to this point, the galvanometer reading ro.se 

 steadily to 110, decreased unsteadily to 10, and then swung 

 unsteadily between 90 and 10 in such way as to indicate 

 sudden and systematic rises and falls of electromotive force, 

 though the flame was not disturbed and the temperature of the 

 lead was constant, and there was nothing apparent which could 

 have caused these changes. At another time the reading being 

 30, not very steady, I removed the flame, the reading fell 



