502 MESSRS, OMOND AND RANKIN ON THE 



of S.W. and W. at sea-level, but of N.W. and N. on Ben Nevis, and a diminution of 

 N.E., E., and S.E. at sea-level, and of E. and S.E. on Ben Nevis, indicating that 

 the cyclonic storms of this month were of sufficient depth to include Ben Nevis, and 

 give winds there only differing slightly from those at sea-level. The December rainfall 

 of Ben Nevis for these six years is about 70 per cent, above the twelfth part of the 

 annual average. It would appear that the cyclonic winds of December, mostly W. and 

 S.W., are largely drained of their moisture in passing over the hills of western Scotland,, 

 although the region of precipitation in them extends to a height so far exceeding the 

 highest of these mountains as to enormously increase the rainfall there.. 



From what has been stated above, it will be evident that no hill station, yet in 

 operation, can be regarded as being in the true upper currents of the atmosphere, 

 and that we are still dependent for our knowledge of their motions on observations 

 of the cirrus clouds. A very careful series of observations on these clouds has been 

 made at Blue Hill Observatory, Massachusetts, by Mr Clayton. In the year 1887 

 over 1500 observations were made, giving the following percentages of frequency in 

 the directions of motion : — S.W., 17 ; W., 40 ; N.W., 26 ; all other directions, 17. In 

 other words, for two-thirds of the time the clouds were moving from the W. or N.W.,. 

 and for half of the remainder from the S.W., leaving only about one-sixth of the total 

 observations for the other five points. These upper currents have the same character 

 as, and are an exaggeration of, the surface winds of Massachusetts, where N. and N.W. 

 are the points of greatest frequency, and the relationship of the two is akin to that 

 between the winds of Santis or Puy de Dome, and their low-level winds as given above. 

 But Massachusetts lies in a belt of high pressure, and is far removed from the North 

 Atlantic depression which influences the relation between the Ben Nevis and sea-level 

 winds in Scotland. The highest station from which data are available is Pikes Peak in 

 Colorado, 14,134 feet, and the percentages of wind directions as given in the Challenger 

 Report, page 165, are as follows :— N., 10 ; N.E., 9 ; E., 1 ; S.E., 2; S., 5 ; S.W., 32 ; 

 W., 20 ; N. W., 18 ; calm, 3. Here we have a marked predominance of S.W. and W. 

 winds, but nothing like the excess from one direction of the cirrus cloud motion 

 indicated by the Blue Hill observations ; it would seem that even at this great height 

 the observers are not wholly in the upper currents, though the percentage of westerly 

 winds is in excess of that at any low-level station in that part of North America. 

 The measurements of the heights of clouds taken at Upsala confirm this (see Nature, 

 vol. xxxvi. p. 321). They show three layers of clouds — Lower, 2000 to 6000 feet, in 

 or near which most hill stations are ; Middle, 12,000 to 15,000, in which Pikes Peak is ; 

 and upper, 20,000 to 27,000, which are the true upper current clouds. 



Any attempt to elucidate the behaviour of the atmospheric currents in cyclonic 

 storms must be based on a careful study of individual cases, and not on general means, 

 such as are dealt with in this paper. For this we have neither had time nor oppor- 

 tunity as yet, but the following points should be borne in mind when studying the 

 development and behaviour of cyclones. 



