High Temperatures. 197 



should a temperature in this region be desired, the melting 

 point of antimony is preferable. 



Zinc. — The melting temperature of zinc can also be deter- 

 mined in an open porcelain crucible provided the same zinc be 

 not too often used — with too much oxide present, the curves 

 were less well-defined. In graphite the form of the time 

 curve and the temperature are the same as in porcelain. The 

 mean melting temperature may be given as 4:19'0 o . 



Lead.— The only noticeable feature in the determination of 

 the melting point of lead is its small melting or "latent" heat, 

 which makes some attention to the regulation of the oven 

 temperature necessary. The mean melting point is 326 - 9°. 



Cadmium. — The melting heat of cadmium is much greater 

 and sharp curves are obtained with little trouble. Its mean 

 value is 321 '7°. 



In Table XX are contained the determinations of the melt- 

 ing and freezing temperature of antimony, aluminium, zinc, 

 lead and cadmium. 



Tables XXI and XXII contain examples of the time curves 

 as they were observed, the numbers representing the thermo- 

 electric force in microvolts as observed from minute to minute. 



Alloys. — In conclusion certain alloys remain to be mentioned 

 whose melting points we investigated by the wire method only. 



The metals were in thin strips cut from sheets of perhaps 

 0*2 to 0'3 mm thickness, prepared especially for the purpose 

 by the Frankfurt Scheideanstalt. 



As the investigation could only be undertaken in an oxidiz- 

 ing atmosphere, alloys of silver and copper gave no definite 

 results whatever. The silver melts and separates out of the 

 partly oxidized copper, leaving it to interrupt the circuit by 

 breaking at a considerably higher temperature. 



The two alloys 95Ag5Au and 90Agl0Au were especially 

 prepared to ascertain whether the addition of a small quantity 

 of gold would prevent the action of oxygen on the melting 

 point of silver. The observed results seem to negative this 

 assumption. 



The temperature scale for the highest melting points is 

 based upon exterpolation from the curve of the normal element 

 T 2 (Table XYI). 



6. Measurement with Ther mo-elements. 



Some ten years ago when the first platinum — platin-rhodium 

 thermo-elements were obtained from the firm of Heraus for 

 the measurement of high temperatures, both the platinum and 

 the rhodium were so imperfectly purified chemically that the 

 duplication of an element with approximately the same thermo- 



