68 DR HUGH ROBERT MILL ON THE 



11. July 1887. — Apparently more influence is exerted in this section by downward 

 conduction from the hot surface layers than is usual. The up-draught at the head of the 

 Basin is less marked, and at the mouth of the Basin the cold deep water seems to be up- 

 welling and creeping across the Plateau under the mass of warm Channel water, which, 

 pouring into the Basin, produces its customary heating effect a little beyond the position 

 where that effect usually begins to be noticed. 



12. August 1887. — Except for its higher temperature, this section closely resembles 

 that for August 1886. In both, the deep water of the West Arran Basin has been raised 

 above that of the central branch at the same depth, but in this the effect of the warm 

 Channel water cannot be so clearly traced, because no observation was taken on the inner 

 edge of the Plateau. The upwelling at Otter is very clearly marked. 



13. September 1887. — This corresponds closely with the section for September 1886. 

 It shows the temperature at the actual maximum observed at the time when the warm 

 salt water of the Channel was producing its maximum effect throughout the Basin. The 

 manner in which the ridge off Loch Eanza abruptly separates the deep water of the 

 western branch from that of the central is exceptionally well brought out. It is interest- 

 ing to note that some less complete sections which were drawn through the East Arran 

 Basin show a marked similarity to the series now being described. Although there is no 

 bar across the deep channel which runs from off Largybeg to off Kilfinan Bay, the upper 

 section of the Basin shows every mark of isolation in the slower response to temperature 

 change, a result due solely to distance from the powerful influence of the Channel water. 



14. December 1887. — This section illustrates best the utter uselessness of isotherms 

 to delineate temperature distributions when the range is small. Here the whole mass of 

 water is almost homothermic at 48°, the isotherms of 47° and 49° only appearing for a 

 short distance at the head of the Basin. The Channel was much warmer than the Arran 

 Basin, and except in Loch Fyne surface-cooling was being propagated directly throughout 

 the mass of water. 



15. January 1888. — Here there is a somewhat unusual distribution of temperature. 

 The surface of the Arran Basin is cold, the cold layer being thinnest at the head where 

 the Otter upwelling raises warmer water, and thickest on the Plateau which it completely 

 covers, cutting off the warmer homothermic water of the Channel from that of the deep 

 parts of the Basin. 



16. February 1888. — The water on the Plateau appears to continue colder than that 

 on either side, the greater part of the Arran Basin being filled with nearly homothermic 

 water of the same temperature as that in the Channel, the process of cooling having 

 evidently gone on unchecked by the usual freshening of the surface water. 



17. March 1888. — Considerable cooling had gone on since February, and the result- 

 ing distribution of temperature is curiously irregular. The temperature on the Plateau 

 was again lower than that on either side, and it had exercised a strong cooling effect on 

 the neighbouring parts of the Arran Basin. At the head, on the Otter slope, there was an 

 area of considerably warmer water, resembling on a smaller scale the arrangement shown 



