1887.] Metals from Low Temperatures. 



301 



fully manipulated during manufacture, and were selected from the 

 author's standard samples, having the chemical composition given m 

 Table I. 



The range of temperature chosen for the observations was from 

 -45° C. to 300° C. 



The experiments were conducted as follows : — For the measure- 

 ments commencing at the low temperature of — 45 °C, the bars 

 (having previously been slowly reduced to the temperature of 0°C, 

 and then gradually cooled to —18° C.) were placed upright in the 

 bath A (see fig.), and immersed in a freezing- mixture of three parts 

 of calcium chloride and two parts of snow, each of these ingredients 

 previous to mixing being maintained in separate jacketed freezing 



Bath A for Temperature of -45° C. Bath B for Temperature of -18° C. 



G-round Flan. 



G-round Plan. 



Scale, f inch = 1 foot. 



tanks at a temperature of —18° C. The vessel A, containing the 

 bars and the calcium chloride freezing-mixture, was further sur- 

 rounded by another compartment holding a quantity of a freezing- 

 mixture of snow and salt at a temperature of — 20 J C. By this means 

 and by constantly renewing the calcium chloride and snow mixture 

 during the experiments, an uniform temperature of — 45° C, as 

 registered by an alcohol thermometer, was maintained for the experi- 

 ments in the cold bath A. 



Much larger cooling tanks of a snow capacity for each charge of 

 8 cwts. were used for the large forgings, and a large cast metal oil- 

 bath having a capacity of about 70 gallons of oil was used for the 

 highest temperature. 



The bars remained thus immersed in the freezing bath whilst 

 their internal temperature was regularly ascertained by another 

 alcohol thermometer placed in a hole in the centre of the test bar C, 



