Motions of Clouds. 441 



of each other ; if this had not been the case, the difference of motion cirro-stratus minus scud would have been 

 the same when obtained ft-om the several individual comparisons and from the final differences. The mean 

 difference of tlie motions of the ciiTO-stratus minus scud from the individual comparisons = + 16°-4, but ob- 

 tained ft'om the two final results (cirro-stratus minus surface), minus (scud minus surface), it is = 21°-7 

 — 13°'9= +7°'8 ; similarly the difference cirrus minus scud from the individual comparisons = + 12°-8, but 

 obtained from the results (cirrus minus sui'face), minus (scud minus surface), it is =31°-9 — 13°-9 =: -f-lS'-O. 

 The difference of the results by the two methods is evidently due to the different observations from which they 

 are obtained ; the law of sign is unaffected, the amount only is different. If wo take the mean of the results 

 for the comparisons of the thi-ee upper currents with the surface current, we find that 



672 results give the mean upper current minus the surface curi'ent, = + 19°'0 



Similarly, if we combine the means for the comparisons of the cirrous and cirro-stratous currents with the scud 

 current, we find that 



323 results give the mean cirro-stratous and cirrovis cuiTent minus the scud cun-ent, = + 15°'2 



If we now examine the relative motions of the mean of the three upper currents, and of the surface current 

 in each quadrant, we obtain the following results : — 



Quadi-ant N. to E., 142 results, mean upper current minus surface current, = + 4°-3 



E. to S., 41 =+25°'0 



S. toW. 353 =+27°-6 



W.toN., 136 =-t-10°-2 



The mean upper current is therefore least positive of the surface cun'ent in the quadrant N. to E., and 

 most positive in the quadrant S. to W. 



If we compare in a similar manner the mean cirro-stratous and ciiTou.s current with the scud current in each 

 quadrant, we have 



Quadi'ant N. to E., 50 results, mean cirro-stratous and cirrous current minus scud current, = — 6°-7 



E. to S., 16 = + 14°-6 



S.toW., 131 =-Hl8°-0 



W.toN., 126 =+2r-0 



In tlie quadrant N. to E. the mean of the highest two currents is negative of the scud current. This is 

 the only case where such difference is found ; the upper current is most positive of the lower current in the 

 quadrant W. to N. 



It appears, then, that the mean upper current always proceeds from a point positive of the direction of the 

 surface current, and that the motion of the mean highest current, with reference to the scud current, follows the 

 same law, with one exception. This result is in accordance with the conclusions from the causes of the oblique 

 motions of the aerial currents. Currents of air proceeding northwards from more southerly positions, retain a 

 portion of the eastward velocity of the places from which they start ; hence the south-easterly winds become more 

 southerly, and the south winds become more south-westerly as they proceed northwards, the extent of the change 

 of direction depending on the greater or less rapidity with which they lose their excess of eastward velocity, and 

 acquire that of the more northerly latitudes over which they move. This loss of eastward velocity will depend 

 upon the proximity of the aerial stratum to the surface of the earth, and therefore the lower currents of air wUl lose 

 more of their eastward velocity than the higher currents, and the upper current of soutlierly winds will become 

 more westerly than the lower currents. If, in considering the cun-ents of air which proceed southwards from 

 more northerly latitudes, we remember that the lower cm-rents, from their proximity to the surface of the earth, 

 acquii'e the gi-eater eastward velocity of the lower latitudes more quickly than the upper currents, it will be evident 

 that the lower current fr-om the north-west will become less northerly than the upper current, and that the 

 lower current from the north will become less easterly than tlie upper current. There appears to be an excep- 

 tion in the latter case : the highest current of air proceeds from a point less easterly than the middle current. 

 It is believed that this anomaly is dvie to the following cause ; — it fi-equently happens that when the lower 

 current of scud is from a north-easterly point there is an upper current of cirrus or cirro-stratus from a north- 

 westerly point : these two currents cannot have had the same origin, and therefore the explanation of the dif- 

 ferences of motions given above cannot apply. 



ilAG. ADD MET. OBS. 1844. 5 T 



