quiredto crush them was, severally, 7,640 lbs., 7,224 lbs. and 7,058 lbs., per square inch of base, which, if compared with 7,978 lbs. on the one-inch cube of the same seasoned wood, shows an apparent diminution of strength in each of the next larger sizes; the average force required to crush the complete parcel of four sets of cubes being 7,475 lbs. The average strength of the unseasoned pieces of the same dimensions only 4,915 lbs. to the square inch of base.
In the experiments (Table XVII.) on a set of fifteen specimens, each 2 × 2 inches, and severally varying only 1 inch in length from 1 to 12 inches, and then by 6 inches, until a length of 30 inches was obtained, it was found that the piece 5 inches in length bore the maximum pressure of 8,820 lbs., or 3-937 tons on the square inch, the resisting power of each of the others being less, while the piece of 24 inches was crushed by about two-thirds the strain of that of 5 inches in length.
From this it would appear that the proper proportion of sectional area to length for this description of timber is as about 4: 5; or, in other words, the superficial area of the base in inches should not be less than four-fifths the length of the pillar or column in inches.
Thus, if it were required to ascertain the scantling for an Oak pillar, 144 inches in length, to ensure its carrying the maximum of strain, we should have inches for the side of that pillar, and according to the ascertained strength of our specimen, 5 inches in length, this would be capable of bearing a weight of 453 tons. Even if we observe the rule, recom¬ mended by many authorities, of only loading to onefourth of the calculated strength, we may still consider it equal to the strain of 113 tons, while a pillar of Riga
