On the Surface Temperature of North Sea and North Atlantic. 19 



true seasonal fluctuation of temperature, and in part of a tendency 

 for the actual body of water present in a given place to change 

 places with another, characterised by very different temperature 

 phenomena. 



Some minor features in this map are not very easy to explain. 

 For instance, there is good evidence that in a large patch of water 

 lying between Spain and the Canaries, the range of temperature 

 is a full degree centigrade greater than that of the waters which 

 surround it ; but the cause of these increased amplitudes is not 

 obvious. 



It will be seen that the lowest range represented anywhere upon 

 our map is 4° C, which is shown as occurring in a large mass of 

 water southward of Iceland, and extending north of Faeroe into the 

 Norwegian Sea. Our data for this region are somewhat scanty, 

 and it is probable that the actual phenomena are much more 

 complicated than as they are represented in our chart. In Dr. 

 Schott's map, a large patch of water is shown in the Atlantic, 

 between Britain and Labrador, with a mean annual range of 5° C. 

 and less, as in our own chart ; but on the other hand Dr. Schott 

 represents much larger instead of smaller amplitudes north of this 

 region, between it and Iceland. He shows, in fact, the isotherms 

 of large amplitude, characteristic of the Newfoundland and Davis 

 Straits region, bending up not only to the westward, but also to the 

 south and south-east of Iceland. After a careful and repeated study 

 of the harmonics which we have derived from the data given in the 

 Danish Year-Book, I feel confident that the distribution shown in 

 our chart (so far as the Danish data extend) is more nearly correct 

 than is Dr. Schott's. 



We may be sure that there is a region characterised by very small 

 amplitudes in the cold current flowing down the East Greenland 

 coast, and also in the cold waters of Davis Straits, especially on its 

 western side. For, for two different reasons, an exceptionally small 

 amplitude is characteristic both of the very hot and of the very 

 cold regions of the ocean. At the Equator, the amplitude is small, 

 for the simple reason that the seasonal change of air-temperature 

 or of solar radiation is also small ; as a matter of fact, in the 

 Equatorial Atlantic, the range of temperature is not more than 2", 

 and is said to be, over a considerable area, less than 1° C. In the 

 polar waters, on the other hand, there is a small annual range of 

 surface temperature, in spite of the fact that the seasonal changes 

 are very marked indeed ; and the obvious reason is that the water 

 is already ice-cold, or nearly so, and is for the most part intermixed 

 with ice. Below its mean temperature it cannot fall much, or it 

 would freeze ; above its mean temperature it has not time to rise 

 far, for the influx of summer heat is mainly employed in melting the 

 floating ice. 



We have now dealt, briefly, with the temperature phenomena 

 of the Xorth Atlantic, in regard (1) to the. mean temperature of the 

 year, and (2) to the mean range or amplitude of the annual seasonal 

 fluctuation. It remains to consider the phase of this sine-wave, 

 from which we derive the corresponding dates of minimal or maximal 

 temperature. But the phase-angle is not so easy to determine with 

 the requisite accuracy as the other co-efficients of the sine-function. 



