AT THE BOTTOM OF THE WATER. 429 



In January, 1860, I made new experiments. I took three smelting kettles 

 of 650 and 670 millimetres diameter, and a wooden tub of 640 millimetres, and 

 filled them with water from the river, which was at -|- 2 degi ees ; the temperature 

 of the atmosphere was — 2 degrees in the daytime ; it sunk to — 5 degrees in the 

 night. The vessels were placed on supports 200 millimetres in height, with a 

 view to having them surrounded with an equal temperature on all sides. The 

 next day, the four vessels were found to be covered with a layer of ice 120 to 

 140 millimetres thick. On the sides the kettles were covered with a layer of 

 ice of 20 millimetres, and from 15 to 20 on the bottom; this stratum of ice was 

 smooth, and without rugosities. The wooden tub had a stratum of only some 3 

 millimetres on the sides, and a few tufts of crystals. On the bottom appeared 

 some isolated films of ice 100 to 110 millimetres in length, 5 to 6 millimetres in 

 breadth, and 1 to 2 millimetres in thickness, furnished on the edges with small 

 projecting points, planted vertically on the larger film, like the teeth of a saw ; 

 these teeth or lateral crystals were 5 to 6 millimetres long, with a breadth of 1 

 to 2 millimetres. 



These experiments, repeated on different occasions, at a temperature of from 

 — 6 to — 7 degrees, always yielded the same result. A/te)- being covered loith 

 a stratum of ice on the surface, the rcssefs were also lined tcith a stratum of ice 

 on the sides and at the hottom, as might have been foreseen ; this ice on the walls 

 of the vesael is of dif event thickness, according to the conductibiJity and radia- 

 tion of the walls. Thus the wooden tub had a thinner layer of ice on its sides 

 than the smelting kettles, and ordinarily only some needles of ice on its bottom. 



I observed that in the kettles these layers of ice Avere proportionably thicker 

 as the external cold was more intense ; that they were always somewhat thicker 

 on the side-walls than on the bottom, and that when once formed these layers, 

 themselves a bad conductor of caloric, operated as isolating walls, and scarcely 

 received any augmentation in point of thickness towards the interior. I have 

 also observed that the surface of the walls, as Avell as the interior surface of the 

 stratum of ice which covered the liquid, was sometimes striated by minute lines, 

 crossing one another at angles of 30, 00, and 120 degrees. Small pools of water 

 of little depth commonly freeze in films which cross one another under the same 

 angles. 



In order to observe well the formation of ice at the bottom of water I have 

 taken smelting vessels, which I have placed in a refrigerating mixture of snow 

 and mai inc salt yielding a cold of — 16 degrees. The circumambient air was 

 at -f i3 and + 15 degrees; the Avater at 0°. Naturally, under these circum- 

 stances, ice was not produced on the surface of the water, but was produced at 

 the bottom of the vessels. The congelation, however, was not always the same ; 

 at one time needles would be formed, which might be seen to increase sensibly 

 ■ util the ascensional force produced by their slight specific gravity had over- 

 come the feeble adhesion of their minute base, when they became detached, and 

 rising, floated on the surface ; at another time the bottom and the walls were 

 very rapidly covered with a thin stratum of smooth ice, marked with the minute 

 lines already mentioned in speaking of the ice formed on the walls of the kettles, 

 or else with a stratum of spongy ice composed of crystals crossing one another, 

 ike the ice at the bottom of rivers. 



It results from these experiments, as might have been foreseen, that ichenever 

 a mass of water is cooled below 0^, and the walls which enclose it are also at a 

 temi^erature of less than 0°, this water will freeze at the bottom as well as at 

 the surface, and that if water usually freezes only at the surface, it is because 

 the bottom of the water and the walls are- for the most ixirt at a temjjcrature 

 above zero. 



In effect, the earth, always at a temperature above 0^, loses its caloric only 

 at the surfice by radiation or by contact with colder bodies. Not only is the 

 earth, which forms the bottom and the banks of rivers and great reservoirs of 

 water, a very bad conductor of caloric, but the water and ice are still worse con- 



