DIURNAL RANGE OF THE BAROMETER IN FINE AND CLOUDY WEATHER. 451 



Costa Rica, and the greatest difference is at the Arctic station of Jan Mayen, an island 

 surrounded by the moisture-laden air of the North Atlantic Ocean ; and it is further 

 noteworthy that the difference between the clear and cloudy days is as distinct at Jan 

 Mayen in December and January, when the sun was all but permanently below the 

 horizon, as at any other season (see page 463). In general, the differences between fine and 

 cloudy days are greater in high than in low latitudes, and in damp as against dry climates, 

 indicating that the effect depends on the thickness of the atmospheric screen through 

 which the sun's rays penetrate, and on the quantity of condensing water-vapour 

 present, or, more strictly, on the relative humidity. The latter, at all events, seems 

 the most probable explanation of the differences between Sodankyla, Bossekop, and 

 Jan Mayen, all within the Arctic circle, but the first in a comparatively dry part of 

 Finland, the second far north on the west coast of Norway, where rain is more abundant, 

 and the last in the ocean to the north of Iceland. 



Nevertheless the general character of the fine and cloudy day curves is the same 

 at all the stations. Starting from the ordinary diurnal barometric curve, which has 

 a minimum in the early morning, a maximum in the forenoon, another minimum in 

 the afternoon, and a second maximum in the evening, we find that by selecting the 

 fine days only, we increase the forenoon maximum and the afternoon minimum, and 

 diminish the evening maximum and the morning minimum ; while on the cloudy days 

 the reverse takes place, the evening maximum and early morning minimum are increased, 

 and the forenoon maximum and afternoon minimum are diminished. At the tropical 

 stations the difference of the fine and cloudy days respectively from the general average 

 only amounts to a few thousandths of an inch, but in the Arctic regions the changes 

 are so great as to alter the character of the curve. It becomes at Jan Mayen a one 

 period curve with its maximum at noon in fine days and at midnight on cloudy, the 

 fine day curves being more regular from month to month than the cloudy. At the 

 two other Arctic stations the fine day maximum is in the forenoon in most of the 

 months, and the time of maximum of the cloudy day curves is more irregular. It 

 must be borne in mind that the monthly values at these Arctic stations are the means 

 of comparatively few days, generally from four to ten in each month, and as the results 

 have not been smoothed in any way they do not give very regular curves. 



At Trieste, Magdeburg, and Fort William in the temperate zone we find the double 

 character of the diurnal curve persisting on both fine and cloudy days ; but while on 

 the general means of the complete months the forenoon and night maxima of pressure 

 are approximately equal in amount, on fine days the former is greatly increased, while 

 the evening maximum is reduced ; indeed, at Magdeburg and Fort William on the 

 annual fine day mean this latter, though still a maximum point, does not reach the 

 mean value for the day. On cloudy days, on the contrary, the evening maximum 

 becomes predominant, and at Fort William the forenoon maximum is depressed below 

 the mean value in the yearly average, but at the other two stations it still appears as 

 a minor maximum above the mean. At Ben Nevis Observatory the ordinary diurnal 



