304 



Heat Dilatation of Metals. 



[Dec. 22, 



results on Table II appear also to indicate another circumstance of 

 metallurgical interest, viz., that the dilatation was generally rather 

 more in the direction of the length of the metallic cylinders than 

 when measured across the diameter, numerous repeated experiments 

 confirmed this. The result appears more marked in the large round 

 forgings of hammered steels and wrought iron than in the case of the 

 rolled bars. It would therefore seem probable that the crystalline 

 particles of the metals suffer slight permanent alteration of form in 

 the direction of their length during the process of rolling or drawing 

 out, sufficient to very slightly affect their relative longitudinal and 

 transverse dilatations. 



Furthermore, the observations of this memoir, conducted at these 

 very low temperatures, experimentally confirm the suggestion of 

 Professor Tait, inasmuch as the coefficients of dilatation were found 

 generally to decrease with the reduced temperature below 0°C. The 

 author also found such to be the case in his observations on the " Heat 

 Dilatation of pure Ice from very low Temperatures." (See ' Roy. 

 Soc. Proc.,' June, 1886, No. 245, p. 544.) 



It may be remarked that many tons of the various freezing 

 mixtures, snow, &c, were required for the experiments. 



Appendix. — Received January 12, 1888. 



I think it would be misleading to use the figures, given in the 

 second column (Table II), of the dilatation from —18° C. to 100° C. 

 for purposes of exact comparison with the other results. The co- 

 efficients for dilatation between the small margin of —45° C. and 

 — 18° 0. could not be accurately inferred from the results recorded in 

 Table B, because the series of experiments from —18° C. to 100° C. 

 were not made consecutively with the other observations. The mole- 

 cular condition of the metals in that series ( — 18° C. to 100° C.) I 

 consider was probably somewhat different. Judging from the whole 

 of the results over the wider ranges of temperature, I do not think 

 that the coefficients for the temperature between — 45° C. and 

 — 18° C, whenever specially determined, will be found to be of a 

 comparative negative character, or vitiate the general conclusions 

 arrived at in this paper. The whole series of observations I believe 

 coincide in establishing the reduction of the coefficients of heat 

 dilatation with reduced temperature. I hope to make further investi- 

 gations at these low temperatures. 



The Society adjourned over the Christmas Recess to Thursda}', 

 January 12th, 1888. 



