462 MR MILNE ON THE DRYING UP OF THE 



Perhaps I may be permitted to dwell for a moment on the remarkable cir- 

 cumstance, that ice should ever begin to form at the bottom of water. It is known 

 that ice is specifically lighter than water ; and that watei^ itself, after it is cooled 

 down to 39° of Fahr., becomes specifically lighter, and rises to the surface. It is, 

 therefore, not easy at first to see how water should, in any case, begin to fi*eeze 

 ft'om the bottom ; but, nevertheless, there are well authenticated cases of ice form- 

 ing at the bottom, even when there was no appearance of any at the surface. 



This is one of the multifarious subjects which has obtained the attention of 

 M. Arago. His views will be found in a paper, the translation of which ap- 

 peared in Professor Jameson's Philosophical Journal for 1833. The phenomena 

 referred to in M. Arago' s paper, occurred in the Rhine and the Aar. The facts 

 were fortunately observed by scientific individuals, who watched the gradual for- 

 mation of the ice in the bed of the river, and ascertained the temperature not 

 only of the air and the ground, but also of the water at the bottom and at the 

 surface of the current. The water was found always to have been cooled down 

 a few degrees below the freezing point ; and the water at the bottom, at the sur- 

 face, and in the middle, always possessed a general uniformity of temperature. 

 M. Arago states, that the phenomenon had been observed never to occur in 

 stagnant water, but always in running streams, and in those places chiefly where 

 the waters flowed over a bottom bristled with stones, weeds, or other rough sub- 

 stances. The explanation of the formation of ice in these circumstances, is suffi- 

 ciently simple. From contact with the air, the upper surface of the water is 

 cooled down to the freezing point, or to any given point below it. If the water 

 is stagnant, these cold particles, from their less specific gravity, remain on the 

 surface ; but the effect of a current is to intermix the particles belonging to the 

 bottom and the surface respectively, in consequence of which, the whole is re- 

 duced to the temperature fitted for congelation. But, it may be asked, why, in 

 some cases, congelation should begin at the bottom and not at the surface of the 

 stream? There are two reasons for this : (1.) From the obstruction which stones 

 and other objects on the bottom give to the current, the velocity is less near the 

 bottom than at the surface, and thus congelation is more possible at the bottom 

 than at the surface ; for water, when in rapid motion, is found not to freeze 

 readily. This, then, is one reason why, in streams, ice should form first at the 

 bottom. (2.) The presence of certain bodies in an aqueous medium, is, in all 

 cases, known to facilitate crystallization ; than which, a better example cannot 

 be given than water saturated with sugar, which will remain in a state of solu- 

 tion till a piece of wood, or even thread or twine, is dropped into it, when in- 

 stantly crystals of sugar-candy are formed round the foreign body. 



Such is, in general terms, the explanation given by M. Arago of the ground- 

 ice, which has been observed in the German rivers, and which may be seen any 

 winter in most of our Scotch rivers. It is the kind of ice referred to in the letter 



