302 



Mr. T. Andrews. Heat Dilatation of [Dec. 22, 



wherein was also placed a little alcohol. When the bars had reached 

 and remained for some time at the registered temperature of —45° C, 

 each was in turn removed and placed on a suitable wooden frame, 

 and its length instantly and carefully measured by telescopic readings 

 from a delicate micro- vernier gauge (deviations of 2 00 °f an i nCQ 

 were perceptible) also supported on a suitable rigid stand. The bars 

 were then replaced for a short time in the freezing-mixture and again 

 removed and their diameter then carefully measured. No perceptible 

 alteration in the temperature of the bars occurred during the very 

 short time occupied in taking the observations, and frequent tests 

 were made to ascertain this. The average of about thirty measure- 

 ments in each case, both longitudinal and transverse, was regarded 

 as fairly accurate. The dimensions of the bars were taken in a 

 similar manner for the temperature from —18° C, substituting in 

 another cold bath, B, a freezing-mixture of snow and salt to obtain 

 this temperature, and using powdered ice and snow for the observa- 

 tions at 0° C. The higher temperature observations were obtained by 

 heating the whole of the bars in a large hot-water bath for the period 

 necessary to insure that their temperature throughout was as re- 

 quired, and the oil bath was used for the temperature of 300° 0. 

 Liability to temperature errors was, as far as possible, carefully 

 guarded against by constant reference and comparison between the 

 bath thermometers and that in the centre of the test bar, and by 

 keeping the bars immersed during sufficiently long periods. 



The hammered metal?? under observation were large forgihgs of the 

 different metals 7 feet 3 inches long, and 5 inches diameter, planed 

 perfectly square at the ends and turned and polished bright. The 

 measurements were taken on the total length of the forgings, as in 

 the case of the rolled metals, to ensure greater accuracy, the experi- 

 ments being conducted in somewhat similar manner : but owing to 

 the greater length of the forgings, a modification of the method was 

 made. One end of the forging was rigidly secured and the expansion 

 ascertained by measuring the diminishing space between the other 

 end of the forging and a fixed point situated a distance from it. The 

 results are recorded on Table II. 



General B.emarJcs. 



It is interesting to notice that the coefficients of dilatation were 

 greater in the case of the " soft " than the " hard " steels, a circum- 

 stance which may be accounted for by a reference to Table I of the 

 analyses, from which it will be seen that the percentage of combined 

 carbon was much lower in the " soft " than in the " hard " steels, the 

 percentage of pure iron was consequently also greater in the "soft" 

 steels, this caused them to be of a greater specific gravity. The 



