December 2, 1910] 



SCIENCE 



777 



eral problem indicated can be approached 

 in two ■ways, (1) the method of analytical 

 theorems checked by a comparatively few 

 observations; (2) the method of numero\is 

 homogeneous observations leading up to 

 general theories. The earlier researches 

 in meteorology, as a rule, pursued the 

 former plan, and the result was in some 

 cases solutions not applicable to nature. 

 My own judgment favored the second and 

 more laborious system as probably leading 

 to better results in the long run. This 

 opinion was based upon five years' experi- 

 ence with Dr. B. A. Gould at the Argen- 

 tine observatory in stellar astronomy, and 

 three years with Professor Simon New- 

 comb in connection with his Planetary 

 Tables. These masters in science always 

 prepared extensive observational data be- 

 fore proceeding to definitive discussions. 

 Accordingly, in view of the chaotic condi- 

 tion of the records of data in the United 

 States, due to changes of the local hours of 

 observing, changes of place of the station 

 instruments, lack of uniformity in the 

 methods of reduction, it was necessarj^ to 

 devote several arduous years to the mere 

 routine work of overhauling the original 

 observations. This included the readjust- 

 ment and discussion of the following data : 

 (1) The magnetic and the electrical data 

 at 30 stations in all parts of the world: (2) 

 the cloud observations for wind vectors at 

 several stations in the United States and 

 the West Indies; (3) the barometric sys- 

 tem of the United States; (4) the tempera- 

 tures and vapor pressures of the United 

 States; (5) the precipitation records for 

 climatology; (6) numerous observations 

 on evaporation in all parts of the United 

 States. In consequence of this work the 

 "Weather Bureau now possesses a funda- 

 mental barometric system, upon which all 

 forecasting since 1900 has been based, be- 

 ing homogeneous long-record monthly and 



annual values of pressure for about 200 

 stations with normals, 1873-1910, to which 

 the pressure variations due to atmospheric 

 circulation can be referred. There is, also, 

 a homogeneous system of temperatures 

 and vapor pressures, with 33-year normals 

 for more than 100 stations, on which the 

 departures in temperature due to the ac- 

 tion of solar rediation on the circulation 

 can be computed, and these have been made 

 the basis of the entire climatological sys- 

 tem of the United States. The summary 

 of the precipitation records for several 

 hundred stations, in 106 sections, is nearly 

 finished, and these data furnish the basis 

 for studies by engineers interested in 

 water resources as applied to power and 

 irrigation. A report on the subject of 

 practical evaporation is now ready for the 

 press, and covers a large amount of new 

 matter of no little interest and importance. 

 The cloud observations have revolutionized 

 the theory of circulation, and the new re- 

 lations discovered between the radiation 

 of solar energy and transported heat by 

 circulation have opened a large field for 

 important progress. The magnetic terres- 

 trial field Avas broken up into its compon- 

 ents, (1) normal field, (2) diurnal dis- 

 turbing field, (3) meridional deflecting 

 field, (4) the field of perturbations and 

 the relations between these fields and the 

 corresponding currents of ions in the at- 

 mosphere, constitutes a most interesting 

 part of the subject of the transformation 

 of the absorbed radiant solar energy in the 

 earth's atmosphere. Simultaneously with 

 the revision of the observations, we have 

 pursued extensive studies in hydrodynam- 

 ics as related to circulation, in thermody- 

 namics as connected with the causes of 

 circulation, and in electricity and magnet- 

 ism as related to radiation. The recent 

 advances in this last subject by Heaviside, 

 Hertz, Lorentz, Thomson and others, and 



