Change of Heat Conductivity. 431 



end, neglecting longitudinal sound-waves ; as also to a liquid 

 filament retarded, if it possessed uniform stiffness in virtue of 

 any internal motion. 

 Belfast, February 1, 1892. 



Since the above was in print Prof. A. G. Greenhill, F.R.S., 

 has sent to the writer a paper, published in the Proc. Camb. 

 Phil. Soc. vol. iv. 1881, written by him for Prof. Asa Gray, 

 on the greatest height of poles, masts, and trees, consistent 

 with stability. The differential equation involved is, in Prof. 

 GreenhilPs paper, solved by the aid of Bessel's functions, and 

 the investigation is extended to the cases of a solid cane, and 

 a paraboloid of revolution, the general form of the solution 

 for certain other solids of revolution being given. The 

 results for a wire (allowing for a slight difference in the value 

 assumed for E) given by Prof. Greenhill are the same as 

 those above. The function IP tabulated in the curve fig. 2 

 appears, from Prof. GreenhilPs paper, to be connected with 

 J n (fcx m ) by the relation 



U' = ^J_J>^). 

 Belfast, March 16, 1892. 



XL VIII. Note on the Change of Heat Conductivity on passing 

 isothevmally from Solid to Liquid. By C Barus.* 



IN an earlier paper f I gave an account of the volume 

 expansion and of the change of thermal capacity of 

 thymol, observed quite through the temperature interval 

 = 0° C. to 0=50° (melting-point), both for the solid and the 

 liquid state. I have since considerably extended these ob- 

 servations. If p be density and c be specific heat, I found for 

 liquid thymol (0°-50°C.), 



1/p = 1-0113/(1 - (-0007600 + 20)0) 

 and c = '4475(1 + -002380); and for solid thymol (0°-50°C), 



l/^ -9631/(1- '('•0002456 + 2 0)0), 

 and c= -3114 (1 + -003020). 



With these data in hand I was able to attack the corre- 

 sponding problem in thermal conductivity. I made use of a 

 somewhat modified form of H. F. Weber's J method, since it 

 is well adapted for measuring small conductivities and meets 

 other requirements of the present problem. If A be the 



* Communicated by the Author. 



t Cf. Proceedings American Acad. xxvi. p. 818 (1892). 

 | Wied. Ann. x. pp. 103,304, 472 (1880). 

 2G2 



