ing toward the center and extending parallel to the face of the disk.
On a tumbler, three inches and a quarter in depth and two inches and a quarter in diameter at the top, placed with its mouth to the wind, the result was the same. The frost-forms pointed toward the center and were parallel to the bottom of the tumbler. It might be worth while to find out by experiments how deep and how wide a vessel would be required to cause them to deviate from this rule.
Fig. 11 exhibits an iron pipe elbow, part of the deposit on which was affected by the rebound from the longer curved side as the wind passed through it. If a straight section of pipe be placed so that the wind may pass through it unobstructed, the deposit is made on the windward end, of the same thickness as the metal; and it appears as though that part of the pipe had been
Fig. 10.
cast in the pattern prevailing in that storm, and whitened. The outer and inner longitudinal surfaces of the pipe are left bare and dry.
Very pretty experiments may be made with apples, chairs, wheels, tin cans, feathers, and other objects too numerous to mention.
Fig. 12 is an apple with a faithful imitation of a chrysanthemum on one side. This was made at a low temperature and was white. The most beautiful blossoms were those made of sleet, at a temperature of twenty-five to thirty degrees above zero. They
