1888.] S. A. Hill — Psychrometer and Oondensing Hygrometer, 377 



Sworykin, with fair ventilation .„, ., '00072 



Doyere, with sling thermometers «, '00069 



Blanford, under thermometer sheds '00083 



Angot -00085 



Chistoni '00085 



Hazen, with artificial ventilation '00068 



In the last column of Table II. the value of this factor derived from 

 each set of observations is given. The extreme values shown in that 

 column, '00124 and "00072, are almost identical with those found by 

 Regnault. The mean for the 11 sets of observations with good venti- 

 lation is '00078, for the 9 with bad ventilation, '00111, and for the 7 

 with slight but insufficient ventilation, '00101. The fact that substanti- 

 ally the same values have been found for A, under similar conditions as 

 regards ventilation, on the two days of observation at Allahabad, the 

 first day being excessively hot and dry and the last a steamy day in the 

 rains, points to the conclusion that to whatever extent this factor may 

 be dependent on ventilation, it is almost if not quite independent of the 

 degree of humidity. When August's formula is used indiscriminately 

 for all conditions of wind movement, as is now done by the Indian 

 Meteorological Department, the effect must be to exaggerate considerably 

 the variations of humidity both in the diurnal and the annual period. 

 Observations made during the nights of the cold season, or on calm days 

 in the rains, give too high a degree of humidity, whilst those made 

 during the prevalence of the hot winds in April and May yield results 

 somewhat too low. 



The observations now published, if not so numerous or so accurate 

 as those furnished by some other observers, are in two senses more 

 extensive than any I have yet seen ; for they not only include obser- 

 vations showing a greater difference between the air temperature and 

 the dew-point than is often observed in any other country, but some of 

 them have been made near sea-level and others at various heights up to 

 nearly 11,000 feet. They thus enable us to determine whether the 

 barometric pressure should be taken into account in reducing psychro- 

 metric observations, as the theory indicates, or whether in accordance 

 with Glaisher's assumption, recently revived and advocated by Professor 

 H. A. Hazen,* the variations of pressure have no influence on the indi- 

 cations of the wet bulb thermometer. Selecting only the observations 

 with good ventilation, we may tabulate the results as follows : — 



* American Journal of Science, vol. xxx, Dec, 1885. 

 49 



