Secttion III., 1908. [ 187 ] Trans. R. S. C. 



XVIII. — Deficient Humidity of the Atmosphere. 

 II. Comparisons of various forms of Hygrometers. 



By T. A. Starkey, M.B., D.P.H., F.R.S.I., 



Professor of Hygiene, 



and Howard T. Barnes, D.Sc, F.R.S.C, 

 Professor of Physics, McGill University, Montreal. 



(Read May 28th, 1908.) 



In a previous paper ^ on Deficient Humidity of the Atmosphere, we 

 gave some comparisons of the wet and dry bulb hygrometer with the 

 Eegnault instrument and chemical estimation — the results showed that 

 the readings by the first were far from satisfactory in a dry atmosphere, 

 — ^the wet and dry bulb instrument always indicated too much moisture. 



In the examples cited in that paper our instrument gave a relative 

 humidity of 45^ in the laboratory where the actual moisture content was 

 only 6^. By causing the air to circulate round the wet bulb, the read- 

 ing was reduced to 36^, but this was so much in error that, together 

 with a large number of other similar erroneous readings, we felt justified 

 for the time being, in condemning the ordinary wet and dry bulb instru- 

 ment for reading indoors during the winter when the air becomes so 

 dry. 



We do not wish to leave the impression tliat we are trying to under- 

 mine the value of an instrument on which most observers rely. Such a 

 procedure is open to criticism as tending to throw discredit upon an old 

 " standby, " without offering something reliable in its stead. This, 

 however, is not our object. 



Our aim is to find out under what conditions precisely the wet and 

 dry bulb instrument gives accurate readings or deductions. 



As will be shown in the course of the paper, the instrument can be 

 manipulated so as to give fairly approximate results, but hitherto no 

 conditions have been attached to its use; the well known tables of 

 Glaisher have always been taken as correct and sufficient without further 

 data, but, as we shall show, results may be obtained differing widely in 

 their nature by simply altering the conditions of the instrument, and 

 of the currents of air around it, etc. 



AYe think that Glaisher must have had some good foundations, 

 consisting of empirical observations, on which the tables were based, and 

 we cannot imagine otherwise than that he checked the correctness of 

 his observations by some system of chemical bygrometry, for after all this 

 is the real and absolute determination; and any instrument which does 

 not give approximate results when compared with the exact chemical 

 estimate, cannot be said to be of much value. 

 ' » Trans. Roy. Soc. Can. 12. 203. (1906). 



