CLIMATIC CONDITIONS OF THE UNITED STATES. 199 



are so many mathematical steps between the actual observations and 

 the finally resulting daily normal means, that so many of the funda- 

 mental properties and characteristics of the data upon which the latter 

 have been based still remain practically unattainable to the student 

 of these important statistics. 



The enormous amount of work represented by the portion of Bulle- 

 tin S that deals with the normal monthly mean temperatures resulted 

 in the elimination, as far as this was possible, of the error-producing 

 effects of variations and alterations in the hours of observation at the 

 different stations throughout the long period of observations, as well 

 as in the reduction of the variously derived daily means to a homo- 

 geneous system. For an account of the ingenious methods employed 

 in this work the reader is referred to Chapter I of Bulletin S ; for not 

 nearly all of the stations are observations for the full 33 years available, 

 and the possible approximate reductions of the means of short-record 

 stations to a 33-year basis were not carried out in Bulletin S, though 

 a method for this sort of reduction is given on page 32. Whether these 

 reductions have been carried out for the temperature data of Bulletin 

 R we are not informed, but it may safely be supposed that the state- 

 ment of the normal daily means of temperature for the 177 stations 

 dealt with in Bulletin R, approach the truth as nearly as was possible, 

 all circumstances being considered, at the time of the preparation of the 

 bulletin. 



It is interesting and worthy of remark here that the letter of trans- 

 mittal accompanying Bulletin S, signed by Willis L. Moore, Chief, 

 sounds a truly prophetic note in this sentence : 



These data, and the normals that have been deduced from them, wiU form the funda- 

 mental basis for future studies on cUmatology and for the investigation of the relations 

 between plant hfe and the thermal and hygrometric conditions that prevail in nature. 



We have chosen 68° F. (20° C.) as our critical normal daily mean 

 temperature in the present connection, and have determined from the 

 tables of Bulletin R the number of days in the year to which are 

 ascribed normal daily means of 68° F. or above. These days are con- 

 secutive in every case, owing to the smoothing process by which the 

 normals have been derived. There are comparatively few stations in 

 the United States without any such days, and for one station, Key 

 West, Florida, this period of what most dwellers in temperate regions 

 would probably term hot days is extended throughout the year. 

 Scarcity of information has no doubt led us into more considerable 

 errors in the western mountainous area than elsewhere. . \ \j^/ 



\\y 



