CLYDE SEA AREA. 69 



in June 1887. Looking at the section generally, the Plateau as a whole seems a centre 

 of cold, on receding from which to the north and the south along the surface — and still 

 more rapidly on deeper planes — the water grows warmer. The rapid cooling since 

 February, the very low temperature of both these months, and the absence of a layer of 

 very cold surface water, point to the great part which must have been played by down- 

 ward convection caused by increase of density by chilling in the surface layers. 



18. April 1888. — Here again, as in the whole series of the winter 1887-88, the coldest 

 water radiates from the inner slope of the Plateau. The difference in temperature is, 

 however, very slight, and the conditions closely resemble those of April 1886, the 

 water being practically homothermic, and the mean temperature almost the same in the 

 two occasions. It represents the typical arrangement of minimum temperature. 



The succeeding temperature trips were not sufficiently systematic to allow of the 

 observations made during them being utilised for drawing sections. 



In order to compare the distribution of temperature with depth in the eastern and 

 western branches of the Arran Basin, the diagram fig. 23, Plate XXVI., shows the Skate 

 Island soundiDg (S) in the centre, with the isotherms connecting it to Inchmarnoch (I) and 

 Brodick, East Arran Basin (B) drawn to the right, and those connecting it to Inchmarnoch 

 and Carradale, West Arran Basin (C) to the left. The profile is approximate only. If 

 the section were folded along the line of the Skate Island sounding, and then folded in 

 the opposite direction along the two Inchmarnoch lines, as shown in the figures at the 

 lower left-hand corner of fig. 23, the run of the isotherms would show the true relations 

 of the two southern branches, while the central part would carry on the lines up 

 the northern branch of the Basin. Fig. 23 shows that, in June 1887, the Eastern 

 Branch was, at the same depths, considerably colder than the Central. It also shows 

 how the included wedge of cold water (see longitudinal section for June) stretches 

 in a rapidly-thinning slice as far as Brodick in the Eastern branch, but does not enter 

 the Western branch at all. It is thus seen to be purely a product of the configuration of 

 the Central branch of the Basin. 



Fig. 24, Plate XXVII., is a similar diagram for September 1887, and is even more 

 instructive, — the sounding off Loch Ranza, on the ridge separating the Western from 

 the Central trough, being inserted. Down to 50 fathoms, or practically to the level of 

 this ridge, there is little difference between the Eastern and Western branches, both being 

 somewhat warmer than the Central, but beneath 50 fathoms the effect of configuration is 

 brought out in a most interesting manner. The temperature of 51° scarcely occurs 

 in the Western branch, and 50° does not enter it at all, although that isotherm runs up 

 the Eastern and Central branches at an average depth of 63 fathoms. Below 80 

 fathoms the temperature in the Eastern branch is below 49°, and this reaches to 

 Inchmarnoch, accounting for the curious patch of cold water in the section for September 

 1887 (No. XIII. Plate XL), but the Central Basin remains warmer throughout, the lowest 

 temperature in the Skate Island Depression being 49°'l. 



