1 1(! Hi(j<'Jow — The Inversion of Temferature Amplitudes 



tomporaturo svsteiu of the Ignited States is really homogeneous 

 for 87 years, and that the data derived from the large cities is 

 indistingnishal)le from those obtained in localities affording 

 more favorable exposure for the meteorological instruments. 

 Siniilarly, the +81 charts available have spread out the history 

 of the mean monthly temi)eraturc variations from the 83-year 

 normal for the period January, 1878, to December, 1909, 

 inclusive, and it is our problem to analyze and discuss this 

 valuable material. 



The Temperature Departures for the tiniled States. 



These charts are available for several purposes. (1) They 

 permit the adjustment of the temperatures of a short record 

 station to the 33-year normal by the following process : Take 

 the means of the short record series of temperatures month 

 by month, and their departures {v, — v^); take the departures 

 by scaling from the charts for the station during the same 

 interval of time {v — z\)', if the sum of the differences, 

 S [(v, — fs) — (v — 'y„)]=0, equals zero, it follows that the mean 

 temperature of the short series is the same as that of the 33- 

 yeai- .normal series, because the departures average the same 

 amount ; if these mean sums are not eqiial it is easy to com- 

 pute a value for t\ which will make them equal, and then 

 i's = v„. The departures {v — v„) are derived from the charts of 

 long record, while those for {v^ — v^) are derived from the 

 station short record, so that it is only necessary to adjust the 

 latter to the former by correcting v^ in such a way as to give 

 the same system of departures for the station record that 

 appears on the charts on the average. 



(2) It is evident that these charts constitute the record for 

 a small area of the earth's surface of the temperature changes 

 going on over continents and oceans in both hemispheres, 

 eastern and western, or northern and southern, such as meteoi-- 

 ologists require for the study of problems in the world- 

 departure of temperature. The Weather Bureau now 

 possesses tables from which correlative variation charts can 

 be constructed for the barometric pressures, the temperatures, 

 the vapor pressures, and the precipitation in the United States, 

 1873 to 1910, and they will be utilized in all important world 

 problems of the atmosphere for many years to come. Only 

 those who have performed the great labor of reconstructing 

 the massive records in meteorology for the United States for 

 this period of 37 yeai-s can appreciate the work that has been 

 accomplished in this respect. It now gives us a very solid 

 foundation for studies of every description in solar and terres- 

 trial physics, anxl it is hoped that similar data will soon be 

 furnished by other well-organized weather services. 



