TRANSACTIONS OF THE SECTIONS. 



51 



K we examine first the numbers for Makei-stoun, we perceive that for weak south 

 winds there is an excess of days when the horizontal intensity was below the mean, 

 while for strong south winds just the reverse was the case ; a simUar opposition is 

 shown for weak and strong north winds. The east and west winds give more 

 consistent results ; but that the opposition in the former cases and the agreement 

 in the latter are independent of the force or direction of the wind, will be evident 

 on examining the corresponding numbera for Singapore. 



As in this discussion many of the days noted may refer to diiferences of hori- 

 zontal force very little above or below the mean, and as these days of slight dif- 

 ference have the same weight in the above tables as days of great difference, we 

 shall avoid what is objectionable in this method by considering the mean of the 

 positive and negative differences of intensity at both places for each wind at 

 Makerstoun ; these are included in the following scheme, where the quantities are 

 in ten-thousandths of the whole horizontal force at the respective places. 



Direction of 



wind at 



Makerstoun, 



1844. 



South . . 

 East . . 

 North . . 

 West .. 



From this it appears that the horizontal force was, on the average, less than the 

 mean for north and south winds at Makerstoun, and greater than the mean for east 

 and west winds. The results for weak and strong winds are also generally contra- 

 dictorj', that for east winds being the only decided exception. 



When we compare the quantities for Makerstoim and Singapore, we find the 

 signs, with one exception, the same, but the amoimts less at the latter than at the 

 former station. This difference is due to the greater effect of magnetic disturbances 

 on the means in the higher latitude. That this is the case may be easily shown in 

 the present instance by omitting in the discussion the three days in 1844 having 

 the greatest difference of daily mean from the corresponding monthly mean (namely 

 March 29, April 17, and November 22). This may be done the more readily, since 

 none of the three days is connected vdth any change in the direction of the wind, 

 which was blowing between south and west. For these two directions, then, the 

 following are the results, omitting the three days noted of greatest distm-bance. 



Direction of 



wind at 

 Makerstoun, 



1844. 



South . . 

 West .. 



Weak winds. Strong winds. 



Makerstoun. Singapore. Makerstoun. Singapore. 



. -0-79 -0-63 -002 -001 



-0-46 -0 26 +0-34 +023 



All forces of wind. 

 Makerstoun. Singapore. 



-0-40 -032 

 +010 +007 



Here the agreement for the two places is so much more marked as to confinn the 

 explanation given as to the cause of the difference between the quantities for the 

 two places. It may be necessaiy to repeat that the winds with which the Singa- 

 pore bifilar is compared are not the winds blowing at Singapore, but those blowing 

 at Makerstoun in Scotland. 



The results, then, are not only opposed to those obtained from the Roman Ob- 

 ser\-atory, but they are such as to prove that the direction of the \^and is unconnected 

 with these variations. It is scarcely necessary to remark that the final quantities 



4* 



