318 M. Becquerel on measuring High Temperatures. 



its base, is of a sombre blue, and becomes less, as it removes 

 from the wick ; the 2d, is the obscure place in the middle of 

 the flame ; the 3d, is the brilliant envelope which covers this 

 last, and which, properly speaking, is the flame ; and the 4<th, 

 which is slightly luminous, and surrounds the flame. 



We first place one of the junctions of the two wires at the 

 superior limit of the blue flame, where the air, still charged 

 with all its oxygen, begins to meet the flame. The deviation 

 is here 22°.50. When the junction is placed in the white 

 part, or in the proper flame, the deviation is 20°, while in the 

 obscure part round the wick, it is only 17°. Now, when we 

 raise the point of junction to 300°, the deviation was 8°, which 

 corresponds to an electro-dynamic force of 12; hence the in- 

 tensities of the current in the three preceding places will be 

 54, 44, 32, which correspond to temperatures of 1350°, 1080°, 

 780°, upon the supposition that if the force 12 is produced by 

 a temperature of 300, the force 48 will be produced by a tem- 

 perature four times as great. The temperature 1350° (2462° 

 Fahr.) is therefore the greatest that a platina wire j of a 

 millimetre in diameter can assume in an alcoholic flame, and 

 it corresponds precisely to the points of the blue zone which 

 touches the brilliant part of the flame, where we know the 

 greatest heat resides. With respect to the temperature 780° 

 (1436° Fahr.) it cannot represent that of the same wire 

 placed in the dark part of the flame, which surrounds the 

 wick, since the wire receives all the heat of the brilliant enve- 

 lope which it traverses; hence it follows that the temperature 

 is much higher than it would be without this. 



In order to confirm the accuracy of the law which I have 

 used to determine the temperature of each of the parts of a 

 flame, or at least of the wires immersed in them, I have ope- 

 rated with wires of platina of any diameter not less than the 

 third of a millimetre, and not containing the same quantity of 

 alloy, and I have always obtained the same results. But if 

 this law were not exact, it would inevitably experience changes 

 in operating with wires which contained more or less alloy. 

 Besides, the results obtained by the above method, compared 

 with those given by other methods, may draw attention to a 

 question so interesting to the arts and sciences. 



