20 Fishery Board for Scotland. 



For the phase-angles are all near together, corresponding to dates 

 of minimal temperature (calculated from the first sine-wave) which 

 vary no more than from about the middle of February to before 

 the middle of March, or an angular difference of less than 30° out 

 of 360° ; and in charting our results, to show the differences of 

 phase from one part to another of the map, we should at least 

 like to group together phases differing from one another by not 

 more than 5°, or five consecutive days in the year. Now it is soon 

 found that, in analysing such series of numbers as we have here 

 to deal with, each series consisting of only twelve (monthly) data, 

 and all the numbers being in the form of more or less rough 

 approximations, we cannot always count on so much accuracy as 5° 

 in the determination of the phase-angle ; our results, from one station 

 to another, are apt to be much less concordant, and to form less 

 smooth and even series, than is the case with our determinations of 

 mean temperature and of mean amplitude. Nevertheless it appears 

 quite possible to give a rough and preliminary chart (Fig. 3) show- 

 ing the main features of the variations of phase over the North 

 Atlantic area ; although it must be remembered that this result 

 has been arrived at by a much more liberal process of " smooth- 

 ing " of the individual values than we have applied, or had 

 any occasion to apply, to the results for mean temperature and 

 amplitude. 



According to our chart, such as it is, there are three regions 

 represented where the phase is most retarded, where, that is to say, 

 the epochs of minimal, or of maximal, temperature fall latest in the 

 year. These three regions are (1) off the African coast, near the 

 Cape Verdes, where the minimum appears to come somewhere about 

 the middle of March ; (2) close around Newfoundland, where it 

 comes in the first week of March ; and (3) off the North of 

 Scotland, where it arrives in the last week of February. There are, 

 correspondingly, three regions of acceleration, in all of which the 

 date of minimum temperature falls at, or even a little before, the 

 middle of February. These are (1) off the eastern coast of the 

 United States, in the region round Cape Hatteras ; (2) in the 

 northern Atlantic, to the south of Iceland : and (3) in the east and 

 south-east of the North Sea. 



Let us now, before proceeding to our twelve monthly charts of 

 the temperatures of the North Sea and the neighbouring waters, look 

 a little more closely at the charts which represent, for these waters, 

 the mean phenomena which we have been discussing for the general 

 surface of the ocean. 



In Fig. 4 we have a chart of the Mean Annual Temperature of 

 our British Coasts and the adjacent seas. We see again the main 

 features which we have spoken of in connection with the eastern 

 Atlantic, that is to say, the inflections which follow the course of the 

 so-called " Gulf Stream " or North Atlantic Current, and also of that 

 comparatively small body of cold water which comes down by the 

 east coast of Iceland. But besides the direct influence of these 

 currents, we now perceive that there is a strong tendency for the 

 mean temperature to be lowered in the neighbourhood of land, as 

 compared with the temperature of the open ocean. All the isotherms 

 bend strongly to the southward as we approach the Scottish, the 



