Psychrometric Foronula. 205 



Theoretical Discussion. 



The natural answer to these arguments is that they provide 

 no proof whatever of the truth of the new theory, since they 

 are only what we should expect from our knowledge of the 

 conditions and theory of the psychrometer. The formuLa 



x = f-AB(^-/') 

 is admittedly only a first approximation, but one which is found 

 in practice to give sufficiently accurate results, provided the 

 value of A is regarded as empirical, i.e., it is to be determined 

 by actual experiment for the special conditions under which 

 the instrument is to be used. It might conceivably be improved 

 in many ways ; one that has been tried is the addition of a 

 term C (/ — /')2. and this is if anything supported by Svensson's 

 results,^ since he found the value of A to depend somewhat 

 on that oi /" — /', being larger for low values of /— /' than for 

 high. But any change in the formula which introduces two 

 arbitrary constants in place of one would, ceteris paribus, be 

 likely to increase the accuracy with which it could represent 

 a set of observations. The test of the real value of such an 

 alteration is its correspondence with an actual physical condi- 

 tion, without which we have no certainty that it will apply 

 even approxmately to any other than the particular observa- 

 ,tions from which it is deduced. In the case in question, we 

 conceive that it has not been proved that the proposed factor 

 1] corresponds to an actual definite physical phenomenon, but 

 on the contrary that our observations disprove it. 



The second fact, that the value of ?/ as obtained from the 

 observations discussed is consistently less than unity, so far 

 from demonstrating the presence of a hygroscopic action, is 

 also to be expected from theory, as will be seen by considera- 

 tion of the assumptions underlying the formula. In the usual 

 theory it is supposed that the air is moving past the bulb in 

 such a way that it becomes completely saturated at the tem- 

 perature /' , but does not receive a sensible amount of heat 

 from the surroundings while it is in contact with the bulb. If 

 either of these conditions is not fulfilled, the result will be the 

 same as a diminution of /. And in this connection it is to be 



1 See also Rizzo, Nuovo Cimento, Oct. 1897, p. 241. 



