REPORT ON ATMOSPHERIC CIRCULATION. 45 



hour, day by day, and season by season, at least as compared with what obtains in dry 

 climates and seasons. In truth, as regards dry climates the diurnal variations in the 

 fall of temperature with height, particularly in the warm months of the year, are so 

 varying and uncertain, that it will probably for ever remain a hopeless problem to 

 reduce a barometric observation made at any particular hour to sea level at places, say 

 1500 feet in height and upwards, with a tolerable approximation to the truth. The 

 reason is, that it is not then possible to deduce from the observations made the 

 approximate mean temperature of the stratum of air between the station and sea level. 

 In constructing daily weather charts, the difficulty is in some degree met by combining 

 with the temperature at the time of observation, the temperature at one or two 

 previous observations. In this work all these difficulties are very greatly reduced, 

 since what are dealt with are only the mean pressures and temperatures of series of 

 years. In drawing the isobars and isothermals, greater weight has been given to the 

 observations made at low than at high stations. 



As respects situation, the least variation in the rate of diminution of temperature 

 with height occurs at places near the sea, and particularly on the windward coasts of 

 land areas, and the rate varies from the normal on advancing into inland climates. At 

 high level stations situated on true peaks, the rate closely approximates to the normal ; 

 but on elevated plateaux the deviation is considerable, and increases with the dryness 

 of the climate and the intensity of solar and terrestrial radiation. 



Now as regards this discussion, observations from such stations as the above may 

 be considered as affording sea level pressures and temperatures sufficiently close to the 

 truth as to warrant the using of them as part of the data from which the isobars and 

 isothermals of the globe may be drawn. 



But it is quite otherwise when we come to deal with observations made at stations 

 situated in deep valleys, such as Gries, Klagenfurt, and Cordova, at which tempera- 

 ture is abnormally lowered when terrestrial radiation is in excess, and abnormally 

 raised when solar radiation is strong. For this reason, not only have those stations 

 been wholly left out in drawing the isobars and isothermals, but also all others known 

 to be in situations more or less similar. Since information is often not supplied 

 regarding the physical configuration of the earth's surface where the station is situated, 

 it was found necessary to resort to an examination of the diurnal range of the 

 barometer, as shown at the observed hours of the station, in order to arrive at some 

 knowledge as to whether the station was situated in the open or in a deep valley. In 

 this way stations were marked as supplying data either altogether unsuitable, or only 

 partially suitable in this discussion. 



It is scarcely necessary to add that observations made at stations in deep valleys, 

 not only mislead in drawing the isobars and isothermals of a country, but they are 

 absolutely useless, and even worse, when used as data contributing to the solution of 



