the eye. Methinks I have seen groups of three in different places arranged almost exactly alike.
Returning I overhauled a muskrat's house by Bidens Brook. For want of other material it was composed of grass flags, and in a great measure (one half) of twigs and sticks, mostly sweet-gale, both dead and alive, and roots, from six inches to two feet in length. These were in fact the principal material of it, and it was a large one, two feet above the ice. I was surprised to find that these sticks, both green and dead, had the greater part of them been gnawed off by the rat, and some were nearly half an inch in diameter. They were cut off not at a right angle, with a smooth cut, but by successive cuts, smooth as with a knife, the twig being at the same time bent down, which produced a sloping, and, so to speak, terraced surface. I did not know before that the muskrat resembled the beaver in this respect also. It was chiefly the sweet-gale thus cut, commonly the top left on two feet long, but sometimes cut off six inches long.
I see, as I skate, reflected from the surface of the ice, flakes of rainbow, somewhat like cobwebs, where the great slopes of the crystallization fall at the right angle, six inches or a foot across, but at so small an angle with the horizon
