136 DR HUGH ROBERT MILL ON THE 



the same in Loch Goil and the Gareloch, and rather less than half a degree higher than 

 the surface layer in Loch Fyne. The temperature of the mass of the water in 

 Loch Goil was on the average half a degree higher than in Loch Fyne. The mean 

 annual temperature of the surface layer in Loch Goil was 1°"5 higher than the air- 

 temperature for the years 1886-87, the same as for Loch Fyne, and about quarter of 

 ;i degree less than for the other divisions. The mass of water in Loch Goil, like that in 

 Loch Fyne, averages 0°"4 higher than the air. 



All the observations show that, proportionally to its depth, temperature changes are 

 more restricted in Loch Goil than in Loch Fyne, with which alone it can be compared, 

 and its isolation from oceanic influences appears to be more complete. 



Loch Strivan. 



Loch Strivan is an example of a loch basin imperfectly shut off from the Arran and 

 Dunoon Basins by the broad and not very shallow Bute Plateau, and connected with 

 the Arran Basin also, so far as superficial water is concerned, by the narrow and tortuous 

 channel of the Kyles of Bute. 



The loch is described in its physical features in Part I. p. 648. It differs from the 

 other loch basins mainly in having no sharply-defmed bar, and in its breadth 

 diminishing uniformly from its mouth to the head. The section along the axis of 

 the loch given in Part I., Plate 8, No. 12, brings out its peculiar form, the deepest 

 channel across the Bute Plateau being 25 fathoms, while the greatest depression of Loch 

 Strivan is only 42. For the steepness of its hill-slopes this loch may be compared to 

 Loch Goil or the upper basin of Loch Fyne, while with respect to its easy communication 

 with the sea the most similar subdivision of the Area is the Central Arran Basin. 



Observations were usually made between Toward and Bogany on the Bute Plateau, 

 frequently in Rothesay Bay, at the topographical mouth of Loch Strivan in mid-channel 

 off Strone Point, at Clapochlar in the deepest water rather more than half way up the 

 loch, and in shallow water close to the head of the loch. The weather was more frequently 

 stormy or wet whilst this division was being examined than in the case of any of the 

 others, its free opening to the south allowing the swell coming up the Channel between 

 Bute and the mainland to run straight up the loch. 



The curves of vertical distribution of temperature need not be considered here in 

 detail. They conformed, as a rule, to the types of the Arran Basin, rarely showing an 

 approach to those of Loch Goil. On several occasions these curves showed high hetero- 

 thermicity, layers of water at different temperatures being sharply superimposed almost 

 without mixture. 



I propose here to deal with the change of temperature of the loch as a whole as 

 deduced from the temperature sections constructed for each trip. An exception may, 

 however, be made with regard to the time-depth diagram at Clapochlar (fig. 5, Plate VI.). 

 1 1 corresponds in depth with Stuckbeg, and shows a restricted circulation in the lower 



