No abstract

The thermal conductivity of air is determined in absolute measure between -183 and 218 �C. by the " thick-wire " variant of the " hot-wire " method. The apparatus consists of a platinum wire 11 6 cm. long and 1.5 mm. in diameter mounted in a 90 per cent. Pt 10 per cent. Ir tube of 7 mm. internal diameter. The temperature variation of the thermal conductivity of air can be represented with considerable accuracy by the quadratic formula ������������� kt=5. 75 x 10-5(l +0.00317t-0.0000021t2) The present paper suggests a value for the thermal conductivity of air at 0 �C. from a consideration of the data which have been obtained by various hot-wire methods.

A form of " hot wire " apparatus is described in which the hot wire is replaced by a thin-walled nickel tube � in. in diameter mounted in a similar nickel tube in, in diameter. The gas under investigation occupies the narrow annular space between the two tubes. It is shown that the apparatus is completely free from convection up to pressures somewhat greater than 1 atm, and that the temperature discontinuity effect is inappreciable down to pressures of a few cm. of mercury. The apparatus is used to test the pressure dependence of the thermal conductivity of several polyatomic gases. The thermal conductivity (at 0 �C.) of the gases carbon dioxide and methane is found to increase slowly with the pressure over the range 10-80 cm. of mercury. It is suggested that this effect may be taken as evidence of the incomplete participation of the energy of vibration of the molecules in the conduction process.

While regular use is being made in industry of the thermal insulating properties of powders and other fibrous or cellular materials, the physical principles underlying the processes of heat transfer through such bodies have not received much attention. The problem, in regard to powders, has been studied by Smoluchowski in 1911, and later by Aberdeen and Laby. Smoluchowski’s investigation included the effect of such factors as the grain size, the kind, and density of the gas on the conductivity of powders of different materials. The experimental procedure followed in his work is unfortunately open to serious objections, and in consequence the data obtained by Smoluchowski are probably not reliable. Further the method, a cooling one, could not be applied to powders having high conductivities. In this laboratory, Aberdeen and Laby have studied the conductivity of the insulating powder silox filled with several gases, as a function of the gas pressure. The powder was contained between two concentric cylinders and the heat transfer was measured across a central portion where the flow was shown to be strictly radial. Platinum thermometry was used so that reliable data are to be expected. Their experiments, however, were limited to a single powder, and, at the suggestion of Professor Laby, a more extensive investigation using powders of different materials and of various grain sizes has been undertaken.

Summary.The thermal conductivity of a number of regular and non-regular single metallic crystals has been studied by previous workers. There appear to be no marked differences between the conductivity of single and polycrystal specimens of the regular metals, except at low temperatures, but further confirmation of this statement by additional accurate investigations seems necessary, more particularly at low temp~ratures, where different observers do not agree . .The methods which have been used to determine the thermal conductivity of metals are briefly discussed; the thermal method appears to be the simplest, but the electrical or Kohlrausch's method is better adapted for very low temperatures, i.e., near 20° K.The authors have used the thermal method for determining the thermal conductivity of a copper crystal over a range of temperatures. The heat losses have been made small by using a high vacuum, and then at the room temperature eliminated by a method the theory of which is given below.