CLYDE SEA AREA. 37 



observations. The latter are used only to check the former. Plotting the rate of change 

 per diem ± with the amount of change as ordinates and time as abscissas, a curve of 

 great interest is produced. This curve (Plate XXIII. , fig. 6) shows that, starting with the 

 spring minimum, the rate of warming is zero, but immediately afterward it increases, at 

 first rapidly, then more gradually, until it reaches a maximum just before the seasonal 

 maximum of temperature. At that moment the rate drops to zero, and cooling com- 

 mences, at first rapidly increasing, then continuing steadily until the eve of the spring 

 minimum, when the rate of cooling drops off and heating begins. For 1886 the 

 maximum monthly rate of change per day was + 0°*104 in August, and — 0°*067 in October. 

 In 1887 it was -0°'077 in February, +0°'120 in July, and -0°-090 in November. 



Thus, while the temperature was rising rapidly at the time of the maximum of air- 

 temperature, the water was gaining heat throughout at the rate of 1° in 8*3 days in 1887, 

 and 1° in 9 '6 days in 1886. When cooling most rapidly in October 1886 it lost 1° in 

 15 days, and in November 1887 it lost 1° in 11 days. Storing neat thus appears to be 

 a more rapid process than parting with it. In the case of water this is probably due to 

 solar radiation heating it throughout from surface to bottom, whereas in cooling, water 

 radiates heat from the surface only, and the slower processes of convection or even con- 

 duction are required to send the heat out of the mass. 



The Great Plateau. 



The Great Plateau is the threshold of the Clyde Sea Area, and over its sill, which 

 reaches to within 21 fathoms of the surface at low tide, all exchanges of water between 

 the Area and the ocean have to pass. The southern edge of the Plateau is relatively 

 steep toward the Channel, but its surface slopes up very gradually to the narrow belt of 

 minimum depth. So gentle is the slope that in considering temperature changes the 

 whole may be looked on as a uniform level, across which masses of water oscillate with 

 the tides between the deep water to north and south. The middle of the northern edge 

 of the Plateau runs on to the shore of Arran. Its western edge dips westward into the 

 North Channel beyond Sanda, runs up the east coast of Cantyre as far as Davaar Island, 

 and in Kilbrannan Sound dips rapidly to the deep water of the West Arran Basin. On 

 the east side Ailsa Craig rises rather to the south of the highest part of the Plateau, and 

 the slopes are uniform to the coast of Galloway, and northward bordering the East Arran 

 Basin in a wide shallow round the shore of Ayr. The deep water of the East Arran 

 Basin lies toward the west, much nearer the coast of Arran than that of Ayr, and there 

 is a comparatively steep gradient off the island of Pladda at the south end of Arran. 

 The water on the Plateau was, in its average condition, just perceptibly less dense than 

 that of the Channel, and the surface and bottom salinity were practically equal. 



Observations were made most frequently on the western part of the Plateau, for in 

 uncertain weather the deep water of the Channel could be reached best to the south of 

 Sanda, and Campbeltown Loch was the only convenient harbour from which to work. 

 Distance from any harbour made the number of observations in the south-eastern part of 



