CLYDE SEA AREA. 87 



(No. 10); 9° in 1 fathom from 1 fathom to 2 fathoms (No. 11), — in this case the 

 range of the fathom above and the fathom below were only 0°'4 each ; 8° "2 in 1 fathom 

 (0-1 of No. 6), but here there was a change of 5° in 4 inches, the surface of the loch 

 being frozen at the time, and the weather dead calm. Under the ice, at 2 inches depth, 

 the temperature was 36°, and at 6 inches 41°, at 6 feet 44°. The other cases are 

 2°"7 in 1 fathom from 1 to 2 fathoms (No. 14) ; 8°*8 in 4 fathoms, or an average of 

 2° '2 per fathom in No. 21. All these are cases of rapid surface heating or cooling. 



An interesting case of possible erroneous deduction is shown in looking at the mean 

 temperatures of the lowest 5 fathoms : — 



Nos. . .12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 29 21 

 Temperature 41-6 42-9 44-2 45-2 48-7 47"2 45-6 45-2 44-3 45-2 457 46-8 48*5 49-4 49-2 48-2 45-6 43-4 44-0 447 4:V9 477 



Date 27.9.86 17.11.86 30.12.86 5.2.87 29.3.87 10.5.87 16.6.87 8.7.87 



Here observations on Sept. 27th, 1886, March 29th and June 16th, 1887, gave exactly 

 the same result ; and if no other observations had been taken it might reasonably be 

 supposed that at the depth of 35 fathoms the temperature was constant, whereas these 

 figures happened to be observed in course of successive stages of heating up, cooling 

 down, and heating up again. Here the apparently impartial distribution of dates — 

 autumn, spring, summer — would be apt to confirm a rash generalisation. The existence 

 of constant temperature at the bottom in the case of Inveraray and Strachur would 

 have been accepted as proved if three observations in 1886 and two in 1887 had been 

 omitted. Frequent cases of such fallacy by coincidence have impressed on me the great 

 caution necessary in uniting points by a curve. 



On November 5th and 7th, 1887, observations were made, on the former day in calm 

 weather, on the latter day with a strong N.E. (down-loch) wind. Both were in the 

 afternoon, when the tide was about the same phase. On November 5th the mean 

 vertical temperature was 49°*8, on the 7th 49°'6. On the 5th (Curve 14) there was low 

 surface temperature (45°"9), rising to 49°"8 at 5 fathoms, stationary at or above 50° from 

 7 fathoms to 24 fathoms, then sinking to 49° '2. On the 7th (Curve 15) the temperature 

 nowhere reached 50°. It was 49°'8 on the surface, 49°"9 at 10 fathoms, and then fell 

 steadily to 49° "2. This showed thorough mixture of the water by the action of wind. 

 The sounding on the 7th was about 1 fathom deeper than on the 5th, and thus the mean 

 temperature of the latter was slightly reduced. There was practically no loss or gain of 

 heat between the two occasions, the difference being merely a redistribution of tempera- 

 ture by mixture. Curves 20 and 21 were observed on August 24th and 25th, 1888 

 (see fig. 37 in Plate XXIX.). Both days show great range of temperature. On 

 the 24th it fell from 57°'l on surface to 52°'9 at 5 fathoms; on the 25th from 

 60°'8 to 52°. On the 24th it fell gradually to 51°'8 at 12 fathoms, sharply to 

 48° at 22 fathoms, and with very exceptional abruptness to 45°"5 at 24 fathoms, 

 then gradually to 44° - 3 at the bottom. On the 25th the fall was slightly irregular, 

 but in the main steady from 52° at 5 fathoms to 45 0> 8 at bottom. The curves 

 crossed twice at 4 fathoms and 23 fathoms. On the second occasion there was marked 



