100 



DR HUGH ROBERT MILL ON THE 



of Row 1. (q.v.) and included three minima and three maxima, but unfortunately neither 

 maxima nor minima were fully mapped out by observations. The mean temperature was 

 throughout somewhat lower than that of Row I. The mean of the 1 6 average vertical 

 temperatures was 47°7, or o, 4 lower than for Row I. Of the 1G soundings, 9 were in the 

 summer and 7 in the winter half-year. Omitting two soundings which were not repre- 

 sented at Row L, the mean for Row II. comes out as 47°"6, or 0°*5 lower than that for 

 Row I., it being thus apparent that the station inside the barrier was distinctty colder 

 than that outside, largely of course on account of the much greater depth. 



The strong tidal stream running into and out of the narrow and shallow mouth of 

 the Gareloch produces a distinct effect in mixing the water vertically, and when a 

 breeze is blowing against the tide, the commotion produced is very considerable. 



No special observations were made as to temperature, but the appearances were 

 exactly similar to those at Otter Spit (see Loch Fyne), which were carefully investigated. 

 Comparison of the various sections will show how those at Row differ from the others. 



Observations between Barreman and Clynder. — A few observations were taken at 

 this point, about midway between Row II. and Shandon, in the form of a cross section, 

 which is of value as showing the relation of axial to lateral vertical temperature curves, 

 and giving some information as to the form of the isothermal sheets. The observations 

 were practically simultaneous, that off Barreman pier in 2 fathoms being made at 13 h 15, 

 the three deep-water soundings at 13 h 30, 14''0, and 14''20, and the others immediately 

 after. They were made on June 7th during a strong easterly breeze, the "Medusa" 

 going against the wind from the lee to the weather shore. 



The vertical curves were defined by very numerous observations, so that a cross 

 section could be drawn with some confidence. 



The data are as follows. Air temperature, 52 0- l : — 



Table XL. — Cross Section of Gareloch at Chjndcr. 



No. and depth .... 



1 (2 fms.) 



2(13fms.) 



3(16fms.) 



4(1 2 fms.) 



5(3?) 



6 (6 fms.) 



7 (4 fms.) 



No. of points .... 



o 



9 



9 



9 



3 



6 



3 



Temperature, surface 



48-3 



48-0 



47-3 



L-7-5 



47-4 



47-0 



47-0 



Temperature, bottom 



is-:; 



46T 



45-9 



46T 



47-1 



46-6 



46-9 



Temperature, mean 



is-:! 



467 



46-4 



46-6 



47-3 



46-8 



46-9 



The diminution of mean temperature is here uniform with the increase of depth. At this 

 point the width of the loch is exactly 1 mile. 



From the curves a Election (fig. 45, Plate XXXI.) was drawn on a large scale, the iso- 

 therms of each quarter of a, degree being represented upon it. On account of the great 

 closeness of the thermometer readings, this could be done with considerable exactness. 

 The result showed a remarkable decrease of temperature to windward, most rapid near 



