660 REPORT — 1901. 



wiuds at that time is reinforced by the seasonal coldness of the more prevalent 

 winds. 



As an example of the results derived from the inquiry it may be mentioned 

 that the minimum of the second order eflectat the end of April may be attributed 

 to the relative frequency of ' cold ' winds and the relative coldness of all winds at 

 that period, while the corresponding minimum at the end of October must be 

 assigned to the relative frequency of 'temperate' wiuds and the comparative 

 coldness of those winds at that time of the year. 



The second order eti'ect is apparent in a single year's observations, and has, with 

 few exceptions, a larger amplitude in the analysis of the temperature curve of a 

 single year than in that of a mean curve of a number of years. The amplitude for 

 a single year may be as much as 3°, or a quarter of the amplitude of the whole 

 annual variation. 



A similar effect is found in the variation of magnitude of the barometric 

 gradient between London and Valencia, and London and Aberdeen. It is 

 probable that this periodic variation in pressure plays some part in causing the 

 similar variation in temperature. 



A similar effect is also found in the temperature variation of the sea water at 

 stations surrounding these islands, and the atmospheric efi'ect is probably con- 

 nected with this. 



4. On the Effect of Sea Temperature upon the Seasonal Variation of Air 

 Temperature of the British Isles. B]j W. N. Shaw, M.A.^ F.R.S} 



The paper describes an attempt to utilise the mode of geometrical composition 

 and resolution of sine curves of the .same period to resolve the principal seasonal 

 variations of temperature at a station into constituents, which may be called the 

 primary solar constituent, and the constituent due to the surroundings of land 

 and sea respectively. 



The analysis of atmospheric temperature shows that there is a considerable 

 lag in the occurrence of the seasonal variations of temperature at coast stations as 

 compared with inland stations, and a still greater lag in the variations of 

 temperature in the sea itself. 



The variation in sea temperature is regarded as a periodic cause of variation 

 of atmospheric temperature at coast stations, the effect of which is periodic in the 

 same period, and may be compounded with the primary solar effect to give the 

 resultant seasonal variation. 



The effects of these curves of equal period may be represented in magnitude by 

 the numerical value of the amplitudes of the first order curves of the respective 

 temperatiu'e variations, and they may be compounded geometrically by means of a 

 triangle whose sides are proportional to these amplitudes, and are inclined at angles 

 corresponding to the relative epochs of the curves. In such a triangle the following 

 elements are known : — 



1. A side proportional to the observed amplitude at the station. 



2. The difference in epoch between the primary solar cause and the resultant, 

 i.e., the angle betvreeu the sides proportional to the amplitudes of the primary 

 solar and of the resultant effects. 



3. The angle between the sides proportional to the marine and the primary 

 solar effect. 



By assuming the primary solar effect to be the same for places in the same 

 latitude it would thus be possible to analyse seasonal variation of temperature 

 at any place into its elements, and an example is given of this analysis in the case 

 of Kew. A point of some interest arising out of this is the lag in the seasons at 

 sea-coast stations, showing that not only the autumn and winter ai-e late at the 

 sea-coast, but also the spring, so that an early spring is to be sought inland. 



' See Proc. Royal Soc. 



