34 DR HUGH ROBERT MILL ON THE 



water of the Plateau and neighbouring parts of the Arran Basin, showing that by no 

 means inconsiderable mixing had taken place. In the deep water outside the Plateau 

 there appears to be little variation in temperature in different places. On September 

 22nd, 1886, from Mr Matheson's yacht "Oimara," the temperature off the Mull of Cantyre 

 was found to be uniformly 54°*5 in 65 fathoms. Off Corsewall Point, on the eastern edge 

 of the Channel, 30 miles S.E. of the Mull, the mean temperature in 46 fathoms was also 

 54°"5, the upper 20 fathoms being at 54°*3, the lower layers warming up to 54 0, 8. On 

 the same occasion, in 70 fathoms off the Maidens on the Irish coast, 20 miles south of 

 the Mull of Cantyre, there was uniform temperature of 55° '1 from surface to bottom. 

 This shows that the Channel type of curve is not confined to the vicinity of the Mull of 

 Cantyre. As soon as the shallow water of the Great Plateau was entered on, the 

 surface temperature fell, the winter condition being in course of formation. 



During May and June 1887 several serial temperature soundings were made by the 

 Fishery cruiser " Vigilant, " # in about 40 fathoms, off the southern end of the Outer 

 Hebrides. The curves expressing these showed, as a rule, a uniform positive slope from 

 surface to bottom, the change of temperature being about 1° F. per 10 fathoms. The 

 regularity of these curves is very remarkable, and it is by no means improbable that the 

 Channel curve is derived from this form by thorough mixture of the water. My ob- 

 servations on H.M.S. "Jackal" to the west of Lewis in July and August 1887 t suggested 

 a different probable cause, already referred to (ante, p. 25). In the Minch T found 

 curves of nearly uniform positive slope similar to those obtained by the " Vigilant " in 

 June (A, fig. 5, Plate XXIII. ), but in the open Atlantic, west of Lewis, the typical form 

 was that shown at B (fig. 5), and, as already explained, I think it probable that the 

 upper layer at nearly constant temperature is driven by the surface drift against our 

 coasts, and so enters the Channel. Between the Inner and Outer Hebrides, this water 

 may be supposed to be embayed and brought under the influence of local heating. 

 The curve B is, I believe, the typical form for the upper layers of water in the open ocean, 

 at least near lee shores. It may be viewed as a triply compound curve, uniting the 

 homothermic, inverted, and paraboloid positive types. 



Seasonal Change of Temperature in the Channel. — A diagram (PI. IV. fig. 7) was 

 constructed in order to show the variation of temperature in the Channel with regard to 

 depth and time. This shows the isotherms as almost exactly parallel, straight, vertical 

 lines, crowded together at the times of heating and cooling, spread rather farther apart 

 at the maxima, and widely spaced at the minima. The diagram, being coloured on the 

 principle already mentioned, shows a series of vertical strips of colour, the summer of 

 1887 showing much deeper tints than that of 1886, but the general order of change is 

 the same for both. A great widening of the isotherms in June, and a corresponding 

 crowding in July and August, 1887, is probably an effect due to incomplete data. 

 Indeed, the diagram is largely hypothetical as regards the spacing of isotherms inter- 



* Eleventh Annual Report of Fishery Board for Scotland for 1892. Part III. 

 t Sixth Annual Report of Fishery Board for 1887. 



