38 



NATURE 



[May 12, 1887 



average depth of less than 15 fathoms at Otter Ferry and 

 Minard Narrows, and of a much longer and deeper basin 

 beyond ; the maximum depth of the latter (80 fathoms) is 

 found off Strachur. Loch Goil, only 7 miles long and 47 

 fathoms deep in the centre, is cut off from the Dunoon 

 Basin by a barrier rising to within 10 fathoms of the sur- 

 face, and is thus exactly similar in its situation to Upper 

 Loch Fyne. The average depth over the whole Clyde sea- 

 area is 31 fathoms, and it contains approximately 150,000 

 million tons of sea-water. The estuary of the River Clyde 

 is both narrow and extremely shallow, and the river does 

 not appear to affect the Firth to such an extent as the 

 Forth does the firth bearing its name. 



The submarine features of the Clyde sea-area are varied 

 and complicated ; and this character is shared by the 

 surface of the intervening land, producing a diversity of 

 mountain, glen, and plain, and corresponding effects of 

 sunshine, cloud, and mist, that lend to the temperature 

 cruises a picturesque charm such as rarely invests physical 

 research. 



The cruises take place at intervals of about 50 days, 

 and each occupies a little more than a week. Observa- 

 tions are repeated at about sixty stations, distributed over 

 the whole area. The temperature is ascertained at the 

 surface, at 5 and 10 fathoms, and at distances of 10 fathoms 

 •down to the bottom. Whenever a considerable difference 

 is noted in the readings of two adjacent thermometers, 

 •observations are repeated at close intervals between them, 

 so that when the curve of vertical distribution of tempera- 

 ture is drawn, points are most numerous where they are 

 most wanted, at the regions of change of curvature. All 

 temperature observations are made with Messrs. Negretti 

 and Zambra's patent standard deep-sea thermometers. 

 These are mounted in the Scottish frame, and are re- 

 versed by the fall of a brass messenger. Three thermo- 

 meters are used on the line at once. The readings may 

 ■be relied upon to one-tenth of a degree Fahrenheit, except 

 when the sea is rough ; then the very lively motion of the 

 Medusa introduces a little uncertainty, on account of the 

 difficulty of reading. A slight correction for change of 

 volume of the detached column of mercury is necessary 

 when the temperature of the water differs more than 5° F, 

 from that of the air ; the air-temperature being observed 

 by the wet-bulb sling-thermometer. 



A slip water-bottle is used on the line along with the 

 thermometers, and samples of water are secured from 

 various depths. 



The entire set of observations made on the Clyde sea- 

 area, up to November 1886, have been published in the 

 last number of the Scottish Meteorological Society's 

 Journal ; and for the purpose of giving a general idea of 

 the main results as yet ascertained, it will suffice to de- 

 scribe the varying seasonal conditions in three typical 

 regions, and then to indicate the general distribution of 

 teinperature in the whole area throiighout the year. 



In the North Chamiel, near the Mull of Cantyre, observa- 

 tions could only be made on five cruises, and of these only 

 two could be extended far enough to reach deep water, 

 that of April 16, when the weather was remarkably fine, 

 and that of September 22, when Mr. Mathieson, of 

 Liverpool, was kind enough to give the use of his large 

 steam-yacht Oitnara for the purpose. The result of all 

 the observations is shown graphically in Fig. i. The 

 •distribution was always uniform from surface to bottom 

 (except for a variation of not more than \° in the 

 superficial layer) ; and, as the accompanying figures show, 

 there was a steady rise of temperature from April to Sep- 

 tember, while by December there had been a marked fall. 

 It is noticeable that in all cases except December the 

 surface water was a little warmer than that beneath ; in 

 December it was a little colder. Temperature : — 



August 12 September 22 December 25 

 000 



52-5 545 48-5 



April 16 

 o 

 42-0 



June 19 



o 



47 4 



The annual range, so far as observations go, appears to 

 be about 12°*5 F. The uniformity of temperature through- 

 out the mass of water continues over the Plateau, but gives 

 place to a slightly different distribution in the deep Arran 

 Basin. 



Off Skate Island eight observations have been made 

 between March 1886 and February 1887, and the curves 

 presenting their results are given in Fig. 2. The actual 

 figures observed for surface and bottom are : — 



March 27 April 19 June 21 Aug. 10 



o o o o 



Surface 41 4 ... 438 ... 48*3 ... 53*6 



Bottom 4i'5 ... 41 3 ... 440 ... 45'6 



Sept. 26 Nov. 16 Dec. 29 Feb. 7 



Surface 547 ... 49-3 ... 46-6 ... 437 



Bottom 47-4 ... 51-1 ... 47-4 ... 44-3 



The range of temperature on the surface thus appears to 

 be I3°'3, and on the bottom 9^8. The maximum surface 



Fig. I. — Channel. 



temperature was observed in September, the maximum 

 bottom temperature in November. The continuous curves 

 (Fig. 2) show the course of heating ; the broken lines that 

 of cooling. They illustrate the development of conditions 

 hinted at in the curves for the Channel. Starting with a 

 practically uniform temperature of 41^*4 in March, the 

 water had heated considerably on the surface, and cooled 

 very slightly at the bottom., by April. From that time it 

 warmed throughout, the surface most rapidly, and a mass 

 of water next the bottom was warmed uniformly. The 

 depth of this mass steadily decreased, until in September 

 there was an unbroken gradient of temperature, falling 

 from surface to bottom. By November the surface had 

 chilled considerably ; but at 24 fathoms the temperature 

 was the same as in September, and below that depth 

 higher ; there being little change from 30 fathoms to 

 the bottom. In succeeding months the fall of tempera- 

 ture has proceeded nearly uniformly, the curve approach- 

 ing the form of a straight line, gradually becoming more 

 nearly perpendicular. It will be noticed that the curves 

 are not in all cases perfectly regular, but the deviations 



