184 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Temp. 
Yalue of lo gy at break. 
376° C. 
•77 
413° C. 
•59 
466° C. 
•50 
485° C. 
•47 
535° C. 
•47 
625° C. 
•36 
695° C. 
•36 
Since a specimen of brass wire heated above 370° C. invariably required 
two separate values of a to accommodate the points on two straight lines, 
and since the crossing-points of these lines occurred after twenty to thirty 
oscillations, several experiments were carried out with such a wire to 
discover if the average period of oscillation during the first fifty oscilla- 
tions differed to any extent from that during the second fifty. Results 
showed that there was invariably a small decrease in the period of the 
latter, the values obtained for the given specimen being : 
Unheated state : Average time of 1st fifty oscillations =4 mins. 35 secs. 
„ „ 2nd „ =4 „ 35 „ 
After heating to 477° C. : Average time of 1st fifty oscillations = 6 ,, 35 „ 
,, ,, 2nd ,, = 6 ,, 31 ,, 
Wires of Rarer Metals. 
Wires of gold, platinum -iridium, platinum, and Assay silver were got 
in the pure state and tested for their conformity to the empirical law. 
Gold. — Length taken = 29 cms. Diameter = 1 mm. The oscillations 
were found to die down fairly rapidly, and, although one line could not 
approximate to all the observed points, the same value of a sufficed. The 
values were : 
A. a = 20 
o 
(M 
II 
e 
PQ 
n = ’80 
n = *90 
5 = 155-6 
5 = 168*4. 
experiment gave : 
a = 25 
a = 25 
n = -80 
n = *90 
5=155-6 
5=174-7. 
In each of these cases the range taken was eighty oscillations, the 
break occurring after 50 had been observed. 
A rather curious result was obtained on heating this wire and testing 
