26 LECTURES TO SCIENCE TEACHERS. 



at these points let us raise lines indicating the amount of 

 heat which escaped in a minute. Suppose it started at 

 50 above the temperature of the air, it would be observed 

 that there was a loss in one minute of a certain amount of 

 heat represented by a line of a certain height, but when it 

 got to 40, the loss of heat in one minute was less, and 

 only extended up to the height of the next line, and so we 

 go on until the loss of heat is nothing. Then, connecting 

 these points we get a curve representing the loss of heat at 

 every temperature of the bar, and that is what we required 

 to reduce the observations. You will notice at every point 

 of this bar a certain amount of heat is being lost and 

 passing through by conduction to the cooler parts. What 

 we want to measure is the quantity of heat which is 

 passing through the bar at any point. Now you will 

 notice that by the time the heat has passed through the 

 bar it is all dissipated before it can reach the end, because 

 the temperature is the same as that of the air, and conse- 

 quently all the heat which passes through any section is 

 dissipated before it reaches the end. Consequently, if we 

 wish to know the quantity of heat which is passing through 

 the bar at this temperature all we have to do is to measure 

 the quantity of heat which is lost with these different 

 temperatures in the course of a minute. Suppose at one 

 point the difference of temperature between the bar and 

 the air is 50, we know what heat is lost from there, and 

 by adding the quantity of heat lost at every point we 

 know that is the quantity passing through the section 

 from which we started. This method, which was employed, 

 involved most laborious calculations, and it was only after 

 many years that these were completed and led to extremely 

 satisfactory results, showing that the conductivity of iron 

 is not constant, but that it diminishes with the tempera- 

 ture, as the temperature rises the conductivity diminishes. 

 This was a very remarkable result, not only because it 

 was different from what had hitherto been believed, and 

 from what Fourier had assumed, but also because of its 

 ramarkable analogy with electricity. The same experi- 

 menter, a great many years before, had pointed out the 

 most remarkable fact that the different metals lie in the 

 same order according to their conductivity for heat as they 

 do with respect to their conductivity for electricity. Those 



