112 DR HUGH ROBERT MILL ON THE 



a point or two north of west. This showed the full action of the northerly component, 

 and taken in conjunction with the conditions of Section II., suggested that in the 

 ( rareloch an east wind acts as if it blew up, a west wind as if it blew down. 



The tide throughout was in the first quarter of ebb, and thus, as in the previous case, 

 the disturbance of the strata was probably in large measure tidal. 



Section IV., August 3rd, 4th, 1886. — The average temperature of the loch was 52°-9, 

 and of the superficial 5 fathoms 53 0, 1. The two landward stations were observed on 

 August 3rd, with the tide about half ebb, and showed nearly uniform temperature, with 

 nearly horizontal isotherms having a slight seaward dip. The two seaward stations 

 observed on the 4th, just after low-water, showed strong landward dip of the isotherms, 

 and the bar showed a very marked effect of separating the colder water outside from the 

 warmer within. Thus at Row I. the temperature was 52°'l at the surface, and 51°*3 

 at the bottom ; at Row II. it was 53° - 3 at the surface, 52° *2 at the bottom. Between 

 Row II. and Shandon the isotherms dipped gently landward, 



The wind was variable, on account of the rapid passage of small cyclones ; on the 3rd 

 there was a light westerly wind, and on the 4th a light south-easterly, which would 

 partly account for the steep dip of isotherms at the mouth. 



Section V., September 24th, 1886. — The tide was in the first quarter of ebb, and the 

 section shows at Shandon an exceptionally well-marked intermediate maximum, the 

 traces of which are distinctly seen at G-aretachhead. It is unfortunate that the observation 

 at Row II. was omitted on this trip, but there are indications that the intermediate 

 maximum did not extend so far. From September 10th to 22nd an anticyclone had 

 kept up light variable breezes ; the 24th, when the observation was made, was dead calm. 

 The salinity conditions were somewhat peculiar, the water at the head on the surface 

 being Salter than at either of the other stations ; that at Shandon was freshest. The 

 average temperature of the loch was 53° '9. 



Here the surface water, cooling rapidly, seems to have floated on the warmer layers 

 below, which preserved the natural summer distribution of cooling downward to the 

 bottom. 



A similar though less marked distribution of temperature occurred a year later. See 

 Section XIII. 



Section VI, November 11th, 1886. — The average temperature was 50°, that of the 

 upper 5 fathoms 49° '5. The distribution was comparatively simple, cooling being in pro- 

 gress from the surface downward. But there was a distinct trace of an intermediate 

 maximum at Shandon, lying very much nearer the bottom than in Section V. There 

 was a general gentle dip of the isotherms seaward from Garelochhead to Row II., but 

 a sharp dip landward from Row I. to Row II., the result being that at Row II. the cold 

 layer was thicker than anywhere else. This may partly be accounted for by the fact that 

 the Row stations were observed about half-an-hour before and the others about half-an- 

 hour after high-water. The tidal streams about high-water are, however, relatively 

 feeble. 



