THE ROYAL ARTILLERY INSTITUTION. 
305 
One-eighteenth of full size. 
The upper edge, or that cooled in air, had contracted 8 inches, or ^th 
its entire length, and slightly increased in thickness; while the lower 
edge, cooled in water, had expanded *875 inch, making a difference between 
the two circumferences of 8*875 inches. The breadth remained unaltered 
(3 inches), and kept perfectly straight. 
Fig. 3. 
Section showing the amount of contraction. One-half the full size. The dotted lines show the 
original form. 
The quality of the iron was afterwards tested by pieces taken from the 
upper and low T er edges, and also from the centre; the fibrous condition had 
remained unchanged, the specific gravity had not altered appreciably, and 
there appeared to be no deterioration in any part of it. 
Experiment 2. Two hollow cylinders of wrought-iron, 12 inches 
diameter and ^ inch thick each, and respectively 9 inches and 6 inches deep, 
were heated to redness, and cooled by half-immersion in cold water twenty 
times; for effects see figs. 4 and 5. 
The 9-inch cylinder did not alter on the upper edge, cooled in air; but 
Fig. 4. 
