TRANSACTIONS OF THE SECTIONS. 55 



not be able to resist the differential pressure from or towards tlie equator which 

 their motion produces. For this reason, the lowest stratum of air, having its 

 velocity relative to the earth kept down by friction, generally moves from the 

 tropical belts of high barometer to the regions of low barometer at the poles and 

 equator. This is the origin of our S.W. winds, and of the prevalent N.W. winds 

 of the Southern oceans, which must be regarded as constituting an undercurrent 

 towards the pole, beneath a topmost cm'rent, also towards the pole, and a middle 

 return current. Between the tropics, on the other hand, the motion thus generated 

 in the lowest stratum of air coincides with the motion due to difference of tern- 

 peratm-e, and this is probably tlie reason why the trade-winds are more constant 

 than the winds of the temperate zones. 



Excess of temperature and moisture in the equatorial regions is unquestionably 

 the prime mover of the winds, as has long been believed ; but the crossing/ of the 

 winds at the tropic, which has often been coupled with it, is a physical im- 

 possibility. 



The tendency of a moving mass of air to swerve to its own right in the northern 

 hemisphere explains the well-established law (Buys Ballot's), that the -wind, in- 

 stead of blowing at right angles to the isobaric lines, and so running down the 

 steepest gi-adient, usually makes an angle of only 20° or 30° with these lines, 

 keeping the region of lower barometer on its left. The rotation of cyclones is an 

 example of this law ; and the pressure which the spirally inflowing sti-eams exert to 

 their own right in virtue of the earth's rotation is the main cause of the excessive 

 central depression. 



Reference was made to Prof. J. Thomson's paper in the British Association Re- 

 port for 1857, and to papers by Mr. Ferrel in the 'American Mathematical Monthly 

 for 1860, and in ' Nature,' July 20, 1871. 



Observations Physiques en Ballon. By M. Janssen. 



Tlie Influence of the Moon on the Bainfall. By W. Pexgellt, F.B.S. 4'C. 



The author commenced by stating that though many of the popular beliefs re- 

 specting "The Moon and the Weather" were no doubt utterly untenable. Sir J. 

 Herschel and M. Arago coucmi-ed in the opinion that, on the whole, the rainfall 

 was somewhat below the general average about the time of fidl moon, and that the 

 fact was ascribable to the effect of the solar heat absorbed by the moon and 

 radiated by her to us. He then proceeded to show that the heat thus received by 

 us must be gi-eatest when, or -very soon after, the moon was full, when she was in 

 perigee, and (in the northern hemisphere) when she had north declination; that 

 the effect of this heat would be a diminution of the rainfall, not during the lunation 

 as a whole, but during a certain portion of it, and therefore an augmentation 

 during some other period ; that the effect would be variable and never considerable ; 

 and that in the northern hemisphere it would be a maximimi when the moon was, 

 at one and the same time, full, in perigee, and in her highest north declination. 

 The paper was illustrated with several tables and diagrams based on rainfall obser- 

 vations made at Torquay diu:ing eighty-seven complete lunations ending with 

 January 19, 1871. 



The following were amongst the conclusions with which the paper closed : — 



No indication of the moon's influence on the rainfall can be detected in the data 

 furnished by an isolated lunation, or by even a few successive lunations. 



Though it may be doubted whether the rainfall statistics of a period shorter than 

 that in which the moon's nodes complete a revolution, or of a solitary locality, 

 would justify general inferences, the data imder discussion appear to indicate that, 

 in the long rim, the moon does somewhat influence the rainfall ; that on the 

 average the dry period of a lunation extends from the flrst day before full moon to 

 the first day before the third quarter, and the wet period from the day of the first 

 quarter to the second day before the full moon ; that the moon's influence on the 

 number of wet days is less marked ; and that the rainfalls are, on the whole, rather 



least hea'vy when the moon has north declination, and when she is in perio-ee all 



indications harmonizing well with physical considerations. 



