108 Professor E. L. Callendar [March 10, 



transpiration, which had been occasionally suggested by previous 

 wr.iers, but have not as yet (so far as the author is aware) been 

 practically investigated as a means of measuring temperature on the 

 absolute scale. The method of effusion consists in observing the 

 resistance to the efflux of gas through a small hole or orifice in a 

 thin plate. In the method of transpiration the gas is made to pass 

 through a fine tube instead of a small orifice, and the resistance to its 

 passage is observed in a similar manner. These methods may be 

 called "resistance-methods" to distinguish them from the ordinary or 

 " bulb-methods " of pyrometry. They are closely analogous to the 

 now familar resistance-method of electrical pyrometry, and possess 

 many of the advantages of that method in point of delicacy and 

 facility of application. One very obvious and material advantage, 

 especially for high temperature work, is the smallness and sensitive- 

 ness of the instrument as compared with the bulb of an ordinary gas- 

 thermometer. But the most important point of difference, which led 

 the writer to the adoption of these methods, is that the measurements 

 are practically unaffected by occlusion or evolution of gas by the 

 material of the tubes. There is a continuous flow of gas through the 

 apparatus. This flow is very large in proportion to any possible 

 leakage, and it is therefore possible to employ platinum tubes with 

 perfect safety. 



The method of effusion may be very simply illustrated by means 

 of a fine hole in the side of a large and thin platinum tube which is 

 heated by an electric current. The current of air is heated in its 

 passage through the tube before it effuses through the orifice. The 

 heated air expands in volume, and the resistance to effusion is 

 increased in proportion to the temperature to which the air is heated. 

 The increase of resistance may be shown by means of a gas-current 

 indicator or " rheoscope," which consists of a delicately suspended 

 vane deflected by a current of gas. A mirror is attached to the vane, 

 and the deflection is measured by the motion of a spot of light reflected 

 on to a scale, exactly as in the case of the mirror galvanometer when 

 used for indicating changes of electrical resistance. As a standard 

 of comparison, to show the changes of temperature of the tuhe, the 

 changes of electrical resistance of the same tube are simultaneously 

 shown by means of a suitable ohmmeter. 



The method of effusion is a beautifully simple method, and gives 

 a nearly uniform scale ; but it has two disadvantages, which it shares 

 with the thermo-electric method of measurement. (1) It necessarily 

 measures temperature at a point, namely at the point of effusion, and 

 cannot be easily arranged to give the mean temperature throughout a 

 space. (2) It is difficult to make the effusion resistance sufficiently 

 large for purposes of accurate measurement. A large resistance 

 moans a very fine hole, and it is not easy to satisfy the theoretical 

 conditions of the problem with sufficieut accuracy and eliminate the 

 effects of viscosity. 



The method of transpiration is more complicated, and does not 



