On the Surface Temperature of North Sea and North Atlantic. 13 



Like most other " waves " with which physical science has to deal, 

 the seasonal wave of temperature is not a simple wave, but a main 

 wave with which other smaller oscilhitions are mixed up ; and we 

 can analyse these, as in the case of a musical note, into waves of 

 half-period, of quarter-period, and so forth. The first of these waves, 

 corresponding to the " octave " above our main oscillation, is in the 

 case of our temperature phenomenon a six-monthly fluctuation. It 

 is a very curious phenomenon, whose cause is not yet well under- 

 stood, and of it we shall liave but little to say in the present paper. 



When we have defined, in regard to the main temperature wave, 

 the mean-value and the half-range, we get, by addition or subtraction 

 of this latter value from the former, to what we may call the Mean 

 Maximum and the Mean Minimum temperatures. They are not 

 strictly the mean maxima or mean minima of the year, for to 

 determine these we should have to take into our calculations the 

 influence of the half-yearly and other higher oscillations of temper- 

 ature. But as a rule even the half-yearly fluctuation is of no great 

 magnitude, and the higher oscillations are very small indeed, so small 

 that we can scarcely say whether they are real phenomena or (as is 

 much more probal»le) are mere appearances due to our imperfect 

 data. In any case, the approximate Mean Maxima and Minima, as 

 we have defined them, — the maxima and minima, that is to say, of 

 the fundamental sine-wave, — are features of considerable interest, 

 whose local variations are often well worth studying. 



Let us now begin, after this short introduction, to consider the 

 main results of our temperature investigations. 



Our object is to understand the temperature phenomena of the 

 North Sea, and of the waters which surround the other coasts of our 

 own islands. But all our coastal waters are but a part of the great 

 system of waters of the Atlantic Ocean, and we cannot properly 

 understand any of our local phenomena without some preliminary 

 knowledge of the ocean temperatures of the whole North Atlantic 

 basin. 



The temperature phenomena of the ocean are closely inter- 

 connected with the system of ocean currents ; and these in turn are 

 correlated with the system of prevailing winds, and are greatly 

 affected by the form itself of the ocean-basin. 



Were the whole globe covered by a shoreless ocean, the distribu- 

 tion of temperature therein would be comparatively simple ; and in 

 the absence of wind it would be very simple indeed. For the main 

 phenomenon would be that of a steady cooling of the surface of the 

 sea from the Equator towards the poles, the temperature varying from 

 a maximum at the one to a minimum at the other, as a simple 

 function of the cosine of the latitude. At the same time, the waters 

 of the ocean would not be stationary ; for the waters cooled in the 

 neighbourhood of the pole would be rendered heavier thereby, and 

 would tend to sink, while the warmer and lighter waters farther 

 south would always tend to float over the colder and heavier layers. 

 There would, in short, be a steady drift to the northward of warm 

 surface waters from the Equator, while the cold waters of the pole 

 would form a steady return-current at the bottom. The obvious 

 effect as regards the surface temperatures would be that at any point 



