CLYDE SEA AREA. 67 



up-draught of the warm bottom layer raising the temperature of the upper strata. At 

 this end of the Basin the last remnant of water as warm as that of the Channel is found, 

 an interesting instance of the effect of configuration. 



7. February 1887. — Here we find that the Channel water has again cooled down 

 below that of the Arran Basin and is chilling it rapidly from the sea, while surface 

 cooling at the upper end of the Basin seems relatively retarded. This effect may 

 perhaps be accounted for by the middle of winter being the rainiest season, and February 

 the month of minimum surface salinity. Hence in February 1887 (and the condition 

 was similar in November 1886) the density gradient offered unusual resistance to down- 

 ward convection, sufficient to restrict the operation of cooling from the surface to down- 

 ward conduction and the chance action of wind. On this account the Channel water, 

 cooling steadily, is once more furthering seasonal change by mixing, in virtue of its 

 higher salinity, with the Arran Basin water. 



8. March 1887. — The minimum has now passed. It was not so low as in 1886, 

 and Loch Fyne did not come to so low a temperature as the Arran Basin, another 

 important difference. Now the Channel has warmed up rather more than the mass of 

 the water in the Basin, on which, however, it seems to have produced little effect. The 

 surface water of the Basin is also warming by surface exchange of heat, and the last 

 remnant of the warm deep layer appears at the head of the Arran Basin as it did in 

 December 1886. 



9. May 1887. — Heating has here proceeded gradually from the surface downward, 

 and from the ocean inward. At the Skate Island sounding and also in Loch Fyne there 

 was evidence of a patch of lower temperature surrounded by warmer water. The 

 appearance suggests that the vertical movements of circulation follow more or less the 

 configuration of the bottom, descending from the Plateau to the deepest part, creeping 

 up the slope at the head of the Basin, and returning seaward in the upper layers. This 

 rotation of the water in a vertical plane, if it exist, would account for a portion of colder 

 water being left in the centre. 



10. June 1887. — Here we see a much higher temperature throughout than in the 

 previous June, and a more marked heterothermicity in the water. The surface stratum 

 of warm water deepens in a very characteristic way when it passes from the practically 

 homothermic Channel to the centre of the West Arran Basin, and then runs away rapidly 

 in the Central Arran Basin, deepening again over Loch Fyne. The peculiar condition 

 of the Plateau on this occasion is treated in detail on p. 24. The equally interesting 

 mass of cold water, resting against the slope at the head of the Arran Basin, and 

 melting away into warm water above, below, and seaward, has been remarked on at p. 

 23. We may look on this wedge of cold water as the dwindling mass attacked equally 

 by the in-draught from the seaward and warmer parts of the Arran Basin from beneath 

 and by the downward conduction of heat from above. From some cause which does 

 not appear, the customary upwelling along the terminal slope does not seem to have 

 occurred, and so the cold mass has not been entirely isolated as in May. 



