[BARNES] ANCHOR-ICE FORMATION 87 



he himself scarcely seems to have had a glimpse of their importance. 

 He mentions an observation by Desmarest. that in a cloudy sky the 

 gronnd-ice accumulates nuiformly, but is interrupted when the sun 

 shines. Now, what he calls a cloudy sky I conceive to be an atmos- 

 phere loaded with hoar-frost, and rendered hazy by its condensation; 

 for I do not thinlc it possible that a genuine cloud can exist in the 

 atmosphere during a keen frost. Here, then, this observer furnishes 

 a fact in perfect accordance with the information on which I proceeded, 

 viz., that the ground-ice is formed only during a hazy state of. the 

 atmosphere, in other words, during a hoar-frost; whilst he tells tis 

 that the process was interrupted when the sky was clear. 



But j\I. Arago quotes a passage from a paper of Mr. Knight, the 

 celebrated botanist, in the 106th vol. of the Phil. Trans., which brings 

 the matter nearer, if not altogether, to a demonstration, though Mr. 

 Knight himself proposes no theory. The passage is as follows : — " In 

 a morniug which succeeded an intensely cold night, the stones in the 

 rocky bed of the river appeared to be covered with frozen matter, which 

 reflected a thread of silvery whiteness, and which, upon examination, 

 1 found to consist of numerous frozen spicula crossing each other in 

 every direction, as in snow, but not having anywhere, except near the 

 shore, assumed the state of firm compact ice. The river was not at 

 this time frozen over in any part, but the temperature of the water 

 was obviously at the freezing-point, for small pieces of ice had every- 

 where formed upon it in its more stagnant parts near the shore; and 

 upon a mill-pond, just above the shallow streams, in the bottom of 

 Avhich I had noticed millions of Utile frozen spicula floating upon the 

 imter. At the end of this mill-pond, the Avater fell over a low weir, 

 and entered a narrow channel, where its course was obstructed by points 

 of rock and large stones. By these, numerous eddies and gyrations 

 were occasioned, which apparently drew the floating spicula under water ; 

 and I found the frozen matter to accumulate much more abundantly 

 upon such parts of the stones as stood most opposed to the current 

 (where that was not very rapid), below the little falls, or very rapid 

 parts of the river."' 



These are by far the most important observations that have been 

 made on the subject of ground-ice. M. Arago is so much struck by 

 them, that, after having concluded his own theory, he says, " It is not 

 certain that the little particles, mentioned by ]\Ir. Knight, do not play 

 an important part in this phenomenon, which I have entirely over- 

 looked." I verily believe they do; only one element is wanting in 

 the catalogue of Mr. Knight's observations to decide my opinion, and 

 that is, the state of the atmosphere during the preceding night. I 



