CLYDE SEA AREA. 151 



difference between the surface temperature of different divisions was at or immediately 

 before the maximum. Thus, in September 1886 the surface water in the Channel was 

 6°*2 warmer than in Loch Goil, and in August 1887 that in the Gareloch was 7° '6 

 warmer than in Loch Fyne. Comparing the years under observation, we find that in 

 1887 the monthly temperature corresponded during the warming period with that two 

 months later in 1886, and the maximum was on the average 2° higher. This was 

 brought about by the spring minimum of 1887 being 1°'6 higher than in 1886 (0°*7 

 higher than in 1888), and by the increment of temperature between April and May being 

 1° greater, the succeeding monthly increments being a trifle less. The fall of temperature 

 in 1887, between September and October, was l 0- 4 greater than in 1886, so that the 

 averages for the succeeding six months in both years were nearly the same. 



The great feature of 1887, as regards surface temperature, was the rapid heating of 

 early summer, leading to a higher surface maximum by bringing it on while the air 

 temperature was still high, so that when the air-temperature began to fall it very 

 speedily stopped farther heating. I have already shown, by reference to the invariable 

 character of the surface temperature curve being cut at its maximum by the air-tempera- 

 ture curve, that the more rapid the rate of heating, the higher is the limiting temperature 

 reached, while the slower the rate, even when much longer continued, the lower is the 

 ultimate maximum. Hence, in the warmest years for surface water, the retardation of 

 the surface after the air maximum is the least, and the greater the retardation the lower 

 is the maximum. The whole may be put tersely in the words, that for all parts of the 

 Clyde Sea Area the surface water continues to heat until its temperature reaches that of 

 the air ; as soon as this occurs surface-cooling sets in. 



Table LXI V. gives the monthly mean temperature of the mass of water in each division 

 calculated from the curves. This is not such a readily comparable datum as the surface 

 temperature, because of the difference in the average depth of the various basins and in 

 their degree of isolation. In its figures we may, however, look for satisfactory evidence 

 as to the influence of configuration on temperature. The minimum mass temperature 

 occurred in April 1886, in March 1887, and in March or April 1888, thus coinciding in 

 time with the minimum of the surface layers. The warmest month for the mass of water 

 was September in the Channel and the Gareloch, October for all the other divisions in 

 1886 ; August in the Gareloch and September in all the other divisions for 1887, this again 

 corresponding to the date of surface maximum. The correspondence with surface 

 changes is similarly shown by the fact that the Gareloch is the warmest division in 

 summer and the coldest in winter; while the Channel is the warmest in winter, and Loch 

 Fyne or Loch Goil usually the coldest in summer. From January to April 1888 Loch 

 Fyne and Loch Goil were the warmest, their temperature being 1° higher than the 

 Channel in February. While the greatest range between the warmest and coldest 

 division as regards surface temperature was 5°"2, and the least 1°'0 ; the greatest range 

 in mass temperatures was as much as 7°"6, and the least only o, 5. At, or immediately 

 after, the minimum the mass of water in the various divisions was most nearly uniform 



