PRESIDKNTIAl, ADDRESS SKCTIOX A. IQ 



The problem which tlie forecaster has to solve every day is the 

 following: Given the isobaric chart of to-day, to construct the 

 isobaric chart of to-morrow. The degree of certitude of the 

 forecasts is, if not equal, at least largely proportional, to the 

 degree of certitude with which this problem can be solved. It 

 must be confessed that we have no solution of this problem. It 

 is sometimes said that meteorology still awaits its Newton. I 

 am afraid that meteorologists have no such ambition. To solve 

 the problem of the weather by analytical methods seems beyond 

 our limitations. The problem is chiefly one of hydrodynamics. A 

 great deal has been done by mathematical physicists in 

 hydrodynamics ; but when it comes to applying their theories to 

 the atmospheric circulation, the problem Ijecomes a singularly 

 complex one. The physicist, in his theoretical investigation, 

 deals with masses subjected to determined conditions of tempera- 

 ture, pressure, humidity, friction, viscosity, etc. If he introduces 

 variations in the conditions, he fixes the law of these variations. 

 He can then deduce the laws of the motion of fluids, laws which 

 he can even verify by laboratory experiments. But when he 

 turns to the atmosphere, he finds that he has to deal with masses 

 out of all proportion with those he can control in his laboratory 

 experiments. He has to deal with masses that are subjected 

 to no very definite conditions of pressure, humidity, temperature 

 or friction, in which, moreover, these conditions change accord- 

 ing to laws about which he knows little or nothing. He has. 

 therefore, to solve what we may consider tbe most difficult pro- 

 blem of mathematical physics. He can only build up the edifice 

 of his theory by sections, and when he has built up his sections, 

 to go on with our metaphor, he finds that to fit them together is 

 worse than any Chinese puzzle, for the sections are not to be 

 juxtaposed since they compenetrate one another; to use the 

 words of one of the leading meteorologists of the day : 



'■ The atmosphere has no simple circulation, cyclonic or anticyclonic. 

 1)Ut is a mass of interlacings of currents.'' 



Although the analytical method has been used with brilliant 

 success ditring the last fifty years or so, it cannot be said that it 

 has been pushed far enough to give us hopes that we will even- 

 tually find the solution of the weather problem in this way. 



We have, therefore, to turn up the inductive method, based 

 principally in this case on statistics and averages, a method which 

 may perhaps help us to find the laws without, however, telling us 

 anything about their nature or about the way in which they are 

 appHed, as the preceding method would do. Statistical methods 

 are subject to a great deal of distrust. " You can prove any- 

 thing with statistics," is sometimes said. But this is not true if 

 statistical methods are properly applied. Statistics are to be 

 applied within limits imposed by the nature of the subject and 

 the nature of the observations, or you may come to absurd con- 

 clusions. It would be easy, by statistics and averages, to prove 



