114 DR HUGH ROBERT MILL ON THE 



about Row, indicating cold up welling at the latter place. The most rapid change of 

 temperature with depth occurred (at Shandon) from 15 to 20 fathoms, where there 

 was a marked crowding of the isotherms. The tide was at the last quarter of ebb, and 

 a very light south-westerly breeze was blowing, the conditions being anticyclonic. 



Section XII., August 6th, 1887. — This section represents the loch at the maximum 

 temperature ever found, and nearly at the greatest range of vertical temperature. The 

 mean was 57°' ( J, and of the upper 5 fathoms 58°'5. The isotherms dipped steadily land- 

 ward, and showed a nearly vertical arrangement at the bar, which separates water of 

 markedly different temperature. The loch was stratified into a thin top layer of water 

 very warm on the surface, a thick well-mixed intermediate stratum, and a thinner bottom 

 layer at Shandon of rapidly diminishing temperature. It is noteworthy that the water 

 from 57° to 5 4° "4 at Shandon was quite cut off by bottom water over 57°*4 from the cooler 

 water outside the bar. Altogether this section is a characteristic example of great surface 

 heating. The weather was calm, and the tide at half flood, the latter probably accounting 

 for the landward dip of the isotherms. 



Section XIII. , September 30th, 1887. — Average temperature, 54° '0, and of upper- 

 layer of 5 fathoms 53°*9. An excellent example of surface cooling, leaving the great 

 mass of water over 54°, with a thin surface zone and a thinner bottom zone (at the ends 

 of the loch only) slightly cooler. The range of temperature is too slight to discuss move- 

 ments, the whole mass of water being fairly well mixed. 



Section XIV., November 29th, 1887. — The mass-temperature had now become 46°*1 

 (the surface layer of five fathoms being 45° "6). The section showed rapid cooling in 

 progress. The upper layer, colder than 46°, was compressed within two fathoms at 

 Shandon, and 1 fathom at Row I., but expanded to 5 fathoms at Row IT., and 8 

 fathoms at the head. This seems to indicate upwelling at Shandon, and down-sinking 

 at Row and Garelochhead, the Row bar keeping out the warmer deep water. 



There was a very light north-easterly breeze, and the tide was in the last half of flood. 



Section XV., February 9th, 1888. — The mean temperature of the section was 43°'7, 

 and that of the surface layer 43° "5, very similar to the conditions shown in Section VII. 

 The temperature was practically uniform, and the two isotherms which appear run 

 parallel to the configuration of the basin, i.e., they indicate a slight upw T elling at both 

 ends. The wind was N.W., nearly down loch, and the tide in the last quarter of 

 flood, or setting up loch ; the former would perhaps account for the slight upwelling 

 at the head, the latter for that at the mouth. 



Section XVI., March 28th, 1888. — The mean temperature was 4 1 0, 9, practically the 

 same as in Section I. The tidal phase was low-water, and a strong north-east wind blew. 

 Homothermic conditions prevailed throughout, showing thorough mixture. The total 

 range of temperature was in fact only 0°*3, the surface being coolest. 



Section XVII, June 7th, 1888. — Gradual warming from above here appears. The 

 mass-temperature had reached 4G°7, and the isotherms steadily dipped landward. An 

 unusual lapse of time separated the observations at Garelochhead and at Row, the tide 



