CLYDE SEA AREA. 117 



the bar, where, in all probability, the isotherms would be found still nearer the surface 

 than at Row L, and the slope to Row II. would be proportionally intensified. 



The six cases of seaward-sloping isotherms occurred in March 1887, April 1886, May 

 1887, June 1886, June 1887, and December 1886. Thus all with one exception occurred 

 during the process of warming up, and all with the same exception at ebb-tide or at low- 

 water. This exceptional case was the only one in which an intermediate minimum 

 temperature was observed. It occurred during a down-loch gale, which might have 

 accelerated ebb-tide, as the Row observations were made just before high-water. The 

 Section for March must also be passed by, as the whole water, inside and outside, was 

 practically at uniform temperature. No. I., April 1886, was observed at \ hour flood, 

 and while the lower isotherms dipped seaward, the upper dipped landward. The 

 weather was calm. Here we observe the upper layer of warmer water steadily thicken- 

 ing toward and beyond the mouth of the loch as the flood-tide was beginning to enter. 

 The May section, at low-water, shows the isotherms practically horizontal across 

 the bar. The section for June 1886 also shows the bar as producing no apparent 

 effect on a down-loch wind distribution ; and the section for June 1887 shows 

 only a slight dip. 



It is unfortunate that practically all the warming-up sections should have been 

 observed with an ebb-tide, and all the cooling-down sections with a flood- tide ; but a 

 consideration of the diagrams makes it seem likely that it is a tidal rather than a 

 seasonal effect which brings out the typical forms of isothermal dip in the case of the 

 Gareloch. On the other hand, it may be that all we can safely say is : During the 

 months of maximum temperature and of rapid cooling the contrast of temperature 

 between the water inside and outside the Gareloch is much greater than it is during the 

 months of minimum temperature and of rapid warming. 



Mean Temperature of Gareloch. — The mean temperature of the mass of water 

 in the Gareloch was calculated in two different ways. As the depression is of 

 comparatively uniform depth, and the several stations rarely differ much in mean 

 temperature, it seemed reasonable to take the average of the vertical temperature 

 means as that of the whole section. This is done in Table XLIIL, cases where 

 one or more observations were omitted being supplied by a probable value. The 

 last column of this table gives the mean temperature calculated from the 

 temperature sections by measuring the areas between adjacent isotherms and making 

 allowance for the volumes of the 10-fathom slices of water. The temperature of 

 the upper slice of 10 fathoms being a, and that of the lower b, the mean temperature 



was calculated by the formula T= - , _, . Details of this measurement are given in 



4 7 & 



Table XLIV. It is remarkable how closely the two estimates correspond, the mean of 

 the nineteen temperature trips being identical when estimated in either way. The 

 greatest deviations were o, 5 in August 1886, 0°-6 in August 1887, and o, 4 in November 

 1887. 



