PROGRESS OF METKOROLO(JV IN 1S80. 223 



obtained by altering' the expres-siou for radiation, so as to make it follow 

 Didong and Petit's law, and assuming /i to be equal to /i„ and each to 

 be inversely proportional to the velocity of the air. Though the theo- 

 retical exi)ressions for ^ and ^ are different, the resulting formula is 

 identical in form with (5), except that the factor -^jfg'- is omitted. 

 A numerical table shows the effect and importance of the missing fac- 

 tor, wliich is moreover shown to be required by some observations of 

 Ferrel's, from which it appears that the factor A of the temperature 

 difference in the typical psychrometer formula, j)=]), — AB [t-t]), where 

 A is supposed to have a known value in making reductions, increases 

 more rapidly with the tem[)erature of the wet bulb than can be ac- 

 counted for by the mere variation of the latent heat. 



(4.) The values of A for ^,=0) are also tabulat''d for a series of veloci- 

 ties; they show considerable variation for small velocities, and also 

 greater variation with a large bulb than with a small one. On this ac- 

 count it is evidently necessary to provide for a constant ventilation of 

 the psychrometer, and to determine the factor A for the specific arrang- 

 ment of the psychrometer by comparison with a dew-point instrument. 

 For small velocities the constants for each psychrometer must be deter- 

 mined as far as possible under the conditions that will obtain when the 

 instrument is subsequently used for observations. 



(5) The use of psj-chrometer tables, founded upon Eegnault's formula, 

 certainly gives on the average, for psychrometers with small bulbs, too 

 small values for the humidity. 



(0) The question as to whether different values for the latent heat of 

 va|)orizatiou should be substituted in formuhe for water-covered bulbs 

 and ice-covered bulbs respectively was raised by Sworykiu upou his 

 finding, in comparisons before mentioned, that the formula for water- 

 covered bulbs gave satisfactor^Mcductions when applied to ice-covered 

 bulbs. It is pointed out (following Ekholm) that practical importance 

 is to be attached to the fact that below the freezing-point the pressure 

 of water vapor has one value over ice, and another over water at the 

 same temperature. The latter is the greater, and should be taken in 

 dew-point experiments, whereas Eegnault's table (used by Sworykin in 

 the comparison) gives the ice- vapor pressure. When account is taken 

 of this, it is shown that Sworykin's observations do not entitle us to 

 abandon the change in the value of A for ice-covered bulbs, which is 

 required on theoretical}grounds. {Quar. Journdl Roy. Meteor. Soc.) 



JMr. S. A. Hill has made a series of comparative observations with the 

 psychrometer and condensing hygrometer under a wide range of con- 

 ditions. He discusses the difficulties and the limits of error in the use 

 of the dew-point apparatus, which he adopts as a standard for compar- 

 ing with the results given by the psychrometric formula of Regnault. 

 His principal results are: The factor A of the psychrometer formula is 

 approximately, if not entirely, independent of the air i)ressure, ;uul also 

 independent of the relative humidity. On an occasion of extreme dry- 



