412 



NATURE 



{March 6, 1879 



ON THE FREEZING OF LAKES » 



IN pursuance of investigations, now extending over 

 sonae years, into the natural conditions of our seas 

 and lakes, I took advantage of the recent frost to make 

 some thermometric observations in fresh-water lakes 

 covered with ice. 



Most of the observations were made with one of 

 Negretti and Zambra's "half-turn" deep sea thermo- 

 meters, which proved to be a useful instrument for this 

 species of inquiry. It was necessary however to fit it 

 with a suitable inverting contrivance, as the apparatus 

 supplied for this purpose by the makers is quite useless. 

 When this thermometer was accidentally disabled, 

 thermometers on the Millar-Casella type with certain 

 improvements introduced by myself were used. 



Observations were made in Linlithgow Loch at different 

 dates, and in Loch Lomond on January 28 and 29.^The 

 results are given in the accompanying tables. 



Table L — Loch Lomond 



TA15LE II. — Linlithgow Loch 



Depth. 

 Feet. 



3 



6 



12 



Bottom 16 



Mud 16 



Bottom i6i 



18 



21 



24 



30 



36 



42 



44 



45 



46 



47 



Mud 



Mud 



Mud 48 



Mean... 



Temperature, Fahren'aeit, at Station. 



No. I. No. 2. No. 3. ; No. 4. 1 No. 6. No. 7 



34-90 — 

 35-25 36-10 

 37-15 36-80 



38-50 



36-95 



37-30 

 37-40 

 3760 



38-60 



39-75 



36-00 

 36-85 

 37-40 

 37-80 



35-90 

 36-30 

 36-80 



36-90 



37-30 

 37-40 

 37-70 

 38-40 



39-85 



— 37-22 37-83 



36-00 

 36-60 

 37-35 



37-35 



37-50 

 37*90 

 38-45 

 39-80 



41-70 



36-00 

 36-80 

 37-50 



37-80 

 37-80 



38-15 

 38-30 

 39-00 

 40-70 



42"00 



(42-05 



142*00 



38-28 



. ' Substance of two papers read before the Royal Society of Edinburgh on 

 January 20 and February 17, 1879. 



Had the freezing of the loch taken place according to 

 the commonly received idea, that is, had its waters been 

 first reduced uniformly throughout its depth to the tem- 

 perature at which the density of water is a maximum, 

 and the surface layer then cooled further until a covering 

 of ice was formed, we should have expected to have 

 found the remains of this uniform temperature on 

 examining the water after a firm coating of ice had 

 formed. Distilled water reaches its maximum density at 

 39-2° Fahr., and I naturally expected to meet with a 

 considerable stratum of water at or near this temperature. 

 Both in Linlithgow Loch and in Loch Lomond there was 

 a tendency to uniformity in the temperature of the water, 

 but in Linlithgow this temperature was approximately 

 37° F. and in Loch Lomond 34° F. A single glance at 

 the curves of the observations in these two lakes shows 

 that they could never have been developed, if at the time 

 of the formation of the first • coating of ice the bulk of 

 the water had been at a uniform temperature of 39*2'^ F. 



In order to explain the existence of this unexpectedly 

 low temperature, I at first imagined that there might be 

 sufficient saline matter dissolved in the water to lower its 

 temperature of maximum density. The presence of fi/e 

 parts of common salt in one thousand parts of water 

 would have sufficed to lower this temperature to about 

 37°, and as the waterof Linlithgow Loch was otherwise 

 excessively foul, it appeared at first sight to be a likely 

 explanation. It was not however verified by experiment. 

 Although possessing a most offensive odour, the water 

 was remarkably free from saline ingredients, and when 

 its change of volume at low temperatures was com- 

 pared with that of distilled water in the same dilatometer 

 no difference could be detected. Seeing then that the 

 temperature of maximum density was the same as that of 

 distilled water, it was evident that, before being covered 

 with ice, the whole of the water had been cooled down 

 much below that temperature, and that this effect had 

 been produced in a still more marked manner in Loch 

 Lomond. 



Let us consider what would be likely to take place during 

 the cooling and freezing of a lalce such as the frozen part 

 of Loch Lomond. The water would be cooled down 

 gradually by radiation from the surface, and we may 

 admit that at some date, probably early in December, 

 the whole water from surface to bottom would have a 

 sensibly uniform temperature of 39-2° F. I believe that 

 even in a very small lake there would be no date when 

 the whole of the water would be uniformly at the tem- 

 perature of maximum density, but it is a condition to 

 which it would approximate in a greater or less degree, 

 according to local circumstances. As the cold con- 

 tinued, circulation would be completely stopped, and 

 cooling would be confined to the surface-layer, supposing 

 the climate of the surface to be absolutely identical all 

 over the surface of the lake. It is impossible that this 

 condition can be fulfilled for more than an instant of 

 time, for it would be disturbed by the slightest^ movement 

 of the atmosphere. As a matter of fact there is great 

 diversity of climates even at points close to each other, 

 causing among other effects great variations in the tem- 

 perature of the surface water. Did lakes cool absolutely 

 uniformly throughout their whole extent, there would be nb 

 reason why ice should begin to form in one part more than 

 another, and the ice would begin to form at one and the 

 same moment all over the lake, which is contrary to 

 experience. 



Let ABC (Fig. i) represent the bed of a lake in section, 

 D E the surface of the water. At the date when the water 

 has approximately the uniform temperature 39-2^ F. 

 D E will represent the isothermal of 39-2. Now let the 

 cooling go on and let the first ice appear, as it naturally 

 would, at the edge. Let E F represent the first piece of ice, 

 which for simplicity's sake we may suppose to have been 

 formed suddenly, and let us consider the effect of its 



