JO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 79 



twice daily, at 8^^ and 2o'\ 75th meridian time, and 

 are set twice daily at S'^ and 2o'\ The extremes given 

 in the summaries are for the civil day, midnight to mid- 

 night, normal * standard time. The monthly means have 

 been obtained by dividing the sum of the mean maxi- 

 mum and mean minimum temperature by 2. 



Precipitation: The rain gages used at the regular Weather Bu- 

 reau stations have a circular catchment area of about 

 8 inches diameter, and the snow, hail, or sleet caught 

 within them is melted and measured as water. The rain 

 gage proper is set within an inclosing cylinder 8 in. in 

 diameter and 2 ft. high, which serves as an overflow 

 attachment in the case of heavy rains and as a snow 

 gage in the winter season. 



The sum total of the depth of rain and melted snow is measured 

 to within o.oi in. at 8'^ and 20'', 75th meridian time, 

 daily. The total precipitation is determined from the 

 amounts recorded daily, midnight t(i midnight, standard 

 of time in local use. 



The snow caught and retained in the gage is melted and measured 

 as water. No correction is applied for the snow that is 

 lost out of the gage by the eddying action of the wind : 

 consequently in some cases the record is less than would 

 be given if the observer had measured cylinders of snow 

 cut from the spots representing the average snowfall 

 on the ground. When it is known that the catch of the 

 snow gage is markedly deficient, an indejiendent ground 

 measurement is made and used as the official record. 

 The loss of both rain and snow, caused by high winds, 

 from gages located on the roofs of tall buildings in 

 which some of the regular stations of the W'eather 

 lUireau are located is undoubtedly larger than is the case 

 at the cooperative stations, where the gages are located 

 in the open country and near the ground, but this loss 

 does not appear to be sufficient to make the monthly and 

 annual sums derived from these two classes of stations 

 wholly inconsistent with each other. 



For a detailed account of the method of reducing the observed 

 barometric pressures the reader is referred to the 

 " Report on the barometry of the United States, Canada, 



* Standard time of the zone in which the station is located. 



