ABSOLUTE THERMAL CONDUCTIVITY OF NICKEL. 375 



Direct Method. 



§ 9. The Apparatus. — The method of determining thermal conductivity by direct 

 measurement of the rate of flow of heat and gradient of temperature is that adopted in 

 the following experiments. This method was used long ago by Clement and by Peclet 

 (vide Ann. de Chimie et de Physique, 3 e torn. ii. p. 107, 1841), and in their hands did 

 not yield satisfactory results as they did not measure the temperatures of the metal 

 itself, but it has been used with success by E. H. Hall, who utilised the metal 

 experimented upon as one of a thermo-electric couple to measure its own gradient of 

 temperature (vide Pro. American Academy, vol. xxxi. p. 271). 



In the present investigation one end of the nickel bar used for Forbes's method was 

 cut off, and an extra thermometer hole was drilled into it. Its surface was repolished. 

 The dimensions were as follows : — 



Diameter, from 4660 to 4 - 667 cm. 



Length, 42 - 55 cm. 



Density, 8'75 grammes per c.cm. 



■vr c rm. 4 tt i Distance in Centimetres from end at which the Rate of 



No. of Thermometer Hole. m « tt a j 



h low of Heat was measured. 



1 2-84 



2 11-20 



3 19-54 



4 27-88 



5 36-17 



A shorter length would have sufficed, and it would have been an advantage to have 

 made more thermometer holes. The bar was fitted up so that one end could be kept at 

 any constant high temperature, while a flow of water could be kept cooling the other, 

 the rise of temperature of the water and the mass of water passing per unit of time 

 being measured. These data were sufficient to measure the rate at which heat left the 

 end of the bar. The gradient of temperature at any point is given by the tangent to 

 the curve drawn from the readings given by the thermometers. 



A slide bench was erected in front of the table carrying the apparatus, and was 

 arranged to carry a telescope which could be raised or lowered in a vertical line, and at 

 the same time moved to and fro along the bench which was placed parallel to the axis 

 of the nickel bar. The thermometers used in the bar were some of Professor Tait's 

 Kew thermometers from the same stock as those used in the Forbes bar, and they were 

 placed so as to hang vertically, this being tested by a plumb line. As the telescope 

 could only move so as to be always horizontal, parallax was avoided. When the tele- 

 scope was adjusted so that one of the thermometers was in focus, all were in focus for 

 that same adjustment, which was never altered. The readings were estimated by eye to 

 the nearest tenth of a degree. 



A diagram representing a vertical section through the axis of the bar is given in fig. 

 2. The heater was the cast-iron pot, J, which was used with the Forbes bar. The bar 

 was fixed into the circular hole (at K) in the side of it with red lead. A Jena glass flask, 



