^4 DR HUGH ROBERT MILL ON THE 



thermic change of temperature in deep water as a sign of free mixture of the water, 

 we are justified in accepting the curious form of the Loch Fyne curves as evidence of 

 the normally restricted circulation in the deep water of that basin. The homothermic 

 agencies being restricted, fuller play is given to the influence of radiation and the contact 

 of warm air or warm water in producing temperature changes. The curves appear to 

 be in large measure conduction curves, convection being reduced to a minimum. The 

 average difference between the density of surface and bottom water at Strachur was 

 0-00212, and at Inveraray '00326, compared with 0-00052 at Skate Island. No fall of 

 temperature would be sufficient to invert the density gradient due to salinity, and thus 

 the conditions in Loch Fyne resemble a vessel of water the temperature of which 

 depends on a layer of oil floating on the surface. This question will be discussed more 

 fully when speaking of the temperature sections of Loch Fyne. 



The characteristic feature of these curves is their tendency to assume a sickle shape. 

 How this form is derived from the homothermic will be explained when dealing with 

 the sections, and, for the present, we may start with Curve 2, which in 1886 showed a 

 much more pronounced intermediate minimum than in 1887. Above 15 fathoms this 

 curve shows rapid heating in progress from the surface downward. Curve 3, after the 

 lapse of 51 days in 1886, 47 in 1887, and 83 in 1888, shows no change of temperature 

 at the bottom, nor in 1886 within 15 fathoms of the bottom, but an increasing rise of 

 temperature as the surface is approached, the curve assuming the form of a paraboloid, 

 significant of the most rapid stage of surface-heating. No. 4, after 45 days in 1886, and 

 89 days in 1887, shows an approximation to the S-shape. In 1886 this curve almost 

 exactly conforms to the " oceanic " type. The lower part is a parabola showing the con- 

 tinued descent of heat, but the bottom fifteen fathoms remain unchanged in temperature. 

 Surface-cooling having set in, the upper layers, losing heat in both directions, have given 

 to the upper part of the curve the form of an inverted parabola. A zone of rapid change 

 of temperature occurs between 25 and 30 fathoms. In No. 4 for 1887 later heating- 

 seems to have turned the inverted parabola into a direct one again. Curve 5 is a very 

 fine example of the negative sickle shape, produced by the rapid loss of heat from 

 the surface, and the more gradual transference of heat downwards, although the bottom 

 temperature was not yet raised. Curve 6 is of similar form, but displaced in a negative 

 direction, except at the bottom, where the temperature now begins to rise. In 1886 

 there was a zone of very rapid change of temperature between 45 and 50 fathoms. The 

 progress of surface-cooling has now made the temperature of the upper layers colder than 

 the remains of the previous cold at the bottom. Curve 7 in 1886 is an approximation 

 to a negative parabola, the bottom water having practically reached its maximum while 

 the surface is at the annual minimum. In 1887 there had been rapid and almost 

 homothermic cooling to the bottom. The transition of Curve 7 to 1 of the following 

 year was in both cases brought about by nearly homothermic cooling. 



It is very interesting to notice that, starting in June 1886 at 44 0, 2, the bottom water 

 required 149 days before any change of temperature occurred ; 79 days more raised the 



