Di'.CEHBEK 23, 1904.] 



SCIENCE. 



877 



thunderstorms and the position of the moon 

 in declination. To Professor Arrhenius 

 is also due the remarkable generalization 

 that the pressure of light emanating 

 from the sun causes alike the streaming 

 away from it of comets' tails, the zodiacal 

 light and the aurora borealis. The rela- 

 tion of sunspot frequency, which has a 

 periodicity of aboi;t eleven years, to atmos- 

 pheric changes on the earth, especially as 

 manifested by barometric pressure, rainfall 

 and temperature in India, has been investi- 

 gated, and the coincidences, even if nothing 

 more, which have been shown to exist by 

 Sir Norman Loekyer and his son are sug- 

 gestive. It may be pointed out that the 

 same action of the sun might cause simul- 

 taneously increased rainfall in India and a 

 deficiency of rainfall in England, because 

 rising currents in one region are necessarily 

 accompanied by descending currents else- 

 where, and, therefore, no objection can be 

 offered to a theory of cosmical influence 

 which produces different weather condi- 

 tions in different parts of the globe. 



Since the sun is the source of our energy, 

 the discovery of any variation in the heat 

 emitted is of the deepest interest and the 

 important investigations of Professor 

 Langiey are now to be supplemented by 

 the broader work of a committee appointed 

 by the National Academy of Sciences and 

 also by an international committee, with 

 the general object of combining and dis- 

 cussing meteorological observations from 

 the point of view of their relation to solar 

 phenomena. It does not seem improbable, 

 therefore, that eventually we may have 

 seasonal predictions of weather, possessing 

 at least the success of those now made daily, 

 and that possibly forecasts of the weather 

 will be hazarded several years in advance. 

 The value of such forecasts, as affecting 

 the crops alone, would be of inestimable 

 benefit to mankind, and predictions already 

 made in India for the ensuing season, while 



not entirely successful, have still proved 

 advantageous. A number of short cycles 

 in the weather have been detected, includ- 

 ing a seven-day period in the temperature, 

 which Mr. Clayton found could be used 

 for forecasting were it not for an unex- 

 plained reversal in the phase of the tem- 

 perature oscillation. 



The interesting question of the value of 

 meteorological observations may appro- 

 priately conclude this address. Professor 

 Schuster, the English physicist, has re- 

 cently denounced the practice" of accumu- 

 lating these observations with no specific 

 purpose. To an extent this criticism is 

 valid in all the sciences, since those ob- 

 servations are most useful when made by 

 or for the person Avho is to utilize them, 

 but although modern meteorology demands 

 special series of observations to solve such 

 problems as the temperature in cyclones 

 and anticyclones, it is sometimes true that 

 long series of observations made with one 

 object in view may subsequently become 

 valuable for quite another purpose. For 

 the study of climate and its possible change, 

 long-continued observations in each country 

 are a necessity, though these might prop- 

 erly be confined to selected stations from 

 whose normals the values for other stations 

 may be computed. Professor Schuster's 

 wish to limit the number of observations 

 implies that the existing series have been 

 inadequately discussed, for the reason that 

 it is easier to find observers than competent 

 investigators. For this unfortunate condi- 

 tion the weather services of most countries 

 are chiefly to blame, because, being burdened 

 with the routine work of collecting clima- 

 tological and synoptic data and formulat- 

 ing and promulgating weather forecasts, 

 which is the public estimate of their entire 

 duty, most government meteorological or- 

 ganizations concentrate their energies and 

 expenditures on these functions and par- 



