WINDS OF BEN NEVIS. 541 



are so liable to cause " radiation frosts." October presents a complete contrast. The 

 S., S.W., and W. are 8 per cent, below the annual mean at sea-level, and 15 per cent, 

 below on Ben Nevis ; and the N., N.E., and E. winds are 13 per cent, above the yearly 

 average at sea-level, and 1 5 per cent, above on Ben Nevis ; showing an increase in the 

 colder currents, slightly greater at high levels, and a diminution in the warmer currents, 

 markedly greater in the upper atmosphere. Comparing for the same six years the 

 monthly average night radiation temperatures — minimum on grass — as given in the 

 Eegistrar- General's tables for all Scotland, we find that in no August did the mean 

 fall below 41°, and in no September below 38°, while in several Octobers it is near 

 the freezing-point, a result quite in harmony with the analysis of the winds of these 

 months given above. The last column in Tables IV. and V. gives the annual variation 

 in calms and variable winds. On Ben Nevis there is a well-marked yearly period 

 of most simple character, the maximum is in June, July, and the minimum in January ; 

 at sea-level, on the contrary, the maximum is in September, and the minimum in April, 

 while there is a curious secondary maximum in December. Thus the frequency of calms 

 on Ben Nevis has the same annual period as the amount of solar radiation, and is of the 

 same character as are the temperature changes, but is always about one month in advance 

 of these. At sea-level, on the contrary, the number of calms seems to be connected with 

 the barometric pressure, but the curve is so irregular that it is not easy to trace the 

 exact connection. The mean monthly velocity of the wind, independent of direction, is 

 greatest in January and least in June on Ben Nevis, and it would seem that the same 

 solar effect which causes the daily and yearly variation in velocity controls the number 

 of calms. 



In order to see how far the rainfall of Scotland is influenced by the winds, a pre- 

 liminary examination has been made of the fall in each month on the east and west 

 coasts. For the east coast the rainfall of each month at the four stations, Marchmont, 

 Barry, Fettercairn, and Gordon Castle, have been tabulated for the six years treated 

 of in this paper ; and for the west coast, the four stations Rothesay, Fort- William, 

 Roshven, and Scourie. These rainfalls are given in the Jour. Scot. Met. Soc, and 

 a few missing cases have been carefully interpolated by comparison with other places 

 near. Taking the average rainfall for each month at each place, and then grouping 

 the four east coast and four west coast places together, we get a fairly representative 

 rainfall for each side of Scotland. The yearly total of the west coast is nearly 

 double that of the east, the numbers being 56 - 27 and 28"29 inches respectively. 

 To make the months comparable inter se, they were reduced to 30 days each ; thus 

 January having 31 days in it one-thirty-first part was subtracted from its rainfall, in 

 February two-twenty-eighths were added, and so on. The average of these twelve months 

 gives the average month, or what each month would have if the rainfall were distributed 

 equally over the year, and the following table shows the departure of each month from 

 this average in percentages, for the east and west coast respectively, heavy type being 

 above the mean, and italic type below it. 



