Prof. Dove's Outlines of a general Theory of the Winds. 231 



propulsion. It is also quite immaterial whether the currents 

 arising in the north or the south are opposed to each other, 

 or whether they are more or less beneath one another, and 

 inclined towards the meridian. For that very reason 1 con- 

 sider the names northerly and southerly currents to be the 

 most conformable to nature, as making their names independ- 

 ent of the variations which the seasons and local causes may 

 produce in their direction. 



The trade-winds and monsoons are a phenomenon so mani- 

 fest, that it is not necessary to prove their existence. It is 

 otherwise with the law of rotation. 



When in the year 1827 I endeavoured to prove the ex- 

 istence of this law by the calculation of 14,600 observations of 

 the barometer, as many of the hygrometer, and 21,900 of the 

 thermometer, which could not be added as mean quantities al- 

 ready determined, but were to be computed one by one, I did not 

 presume that it would be objected to the result of so laborious 

 an investigation, that of three seamen who had been questioned, 

 one should know nothing of it. The possibility that this could 

 happen, proves more distinctly than the silence of the physical 

 manuals upon it, that philosophers did not acknowledge a 

 law in the variations of the direction of the wind. If, how- 

 ever, we consider the remarkable regularity with which this 

 law is manifested in the variations of the barometer, thermo- 

 meter, and hygrometer in Paris and London, calculated by 

 me not only upon a yearly average, but even in every single 

 month, as well as its perfect independence of the daily period, — 

 results which were confirmed by the interesting memoir of 

 M. Galle relative to Danzig, — we might indeed expect, from 

 the exactness of earlier observers, that at least the direct per- 

 ception of that regularity did not escape them. In a review 

 of older and later writings I have also found many proofs of 

 this perception, but which always remained unnoticed for want 

 of a strict proof. This proof, however, could only be given 

 by passing from the computation of the averages to the com- 

 putation of the mean variations. But the rule generally ac- 

 knowledged as just, that we must commence from the average 

 in the investigation of atmospheric phenomena, has unfor- 

 tunately been understood to mean that we must not go beyond 

 the average in these investigations. 



While I have recourse to direct observations in order to 

 prove the general validity of the law of rotation, I premise, 

 that I myself consider this proof imperfect. The calculation 

 of the observations of the barometer at a single place in 

 North America, and in the interior of Russia, worked out as 

 I have done for Paris and London, would be a more con- 



