THE DIURNAL RANGE OF WIND DIRECTION ON BEN NEYIS. 707 



The average monthly frequency of occurrence of each wind was given in the last 

 volume of observations (see Trans. Roy. Soc. Edin., vol. xliii. p. 501). These were 

 averages for the whole day. The tables appended to this paper have been computed to 

 show the variation at each hour of the day from these general means for each month of 

 the year. The winds were tabulated separately for each hour of the day of each month 

 during the twenty years 1884 to 1903 inclusive, and at the same time reduced to eight 

 points (N., N.E., E., etc.) by dividing the number of entries at the intermediate points 

 equally between the two adjacent points : thus one -half of the entries under N.N.E. 

 were counted to N., and one-half to N.E., and so on ; so giving for each month a table 

 showing the frequency of occurrence of the eight principal wind directions, and of calms, 

 at each hour of the day. From these data Table I. has been prepared. At the foot of 

 each monthly table are the percentage frequencies of each wind for the general mean of 

 the month, and in the body of the table the amount that each hour differs from these 

 general means. Thus in January, for the whole month there are 20 per cent, of N. 

 winds ; at I o'clock there is 1 per cent, more, that is 21 per cent. In the same month 

 there are 9 per cent, of N.E. winds for the whole month, but at 1 o'clock there are 2 

 more, that is 11 per cent. At 12 o'clock there is 1 less of N. wind, making it 19 per 

 cent., and 1 less of N.E. wind, making it 8 per cent.; and so throughout the table. 

 Mr James Miller, formerly one of the observers at Ben Nevis, has assisted me in the 

 preparation of these statistics. 



An examination of Table I. shows that there is a maximum of northerly winds at 

 night and during the early morning, and a maximum of southerly winds in the middle 

 of the day. Both these variations are most evident in summer, and almost vanish in 

 the depth of winter. Calms also have a daily period, being most frequent in the 

 afternoon, and at a minimum in the early morning. It is difficult, however, to trace 

 any law in the hourly changes of the other directions of wind in Table I., though west 

 winds resemble north, and east winds south to a certain extent. When the differences 

 contained in the body of the table are added to or subtracted from the mean values at 

 the foot, we obtain the actual percentage frequency of each wind at each hour ; but 

 when the mean frequency is small, it requires a relatively large change in frequency to 

 make a difference of even 1 per cent, at any hour. Thus the less frequent winds are 

 handicapped in comparison with those with a larger mean number of occurrences, in 

 respect to indications of changes from hour to hour. If the mean frequency be 6 per 

 cent., a change of 1 per cent, at any hour is relatively three times as great a change as 

 1 per cent, in a wind whose mean frequency is 1 8 per cent. To get rid of this dis- 

 crepancy, the differences of each wind at each hour from the mean frequency of that 

 wind have been computed as percentage changes of that mean, and thus the proportional 

 change equalised for each wind. That is, the mean frequency, whether large or small, is 

 called 100, and the difference at each hour expressed as a fraction of this. In Table II. 

 the wind frequencies are grouped in seasons of three months each, and the diurnal 

 changes expressed in this manner. At the foot of each section of the table is given 



