Shot were constructed in 1851 to try the effect of rotation in the above-mentioned and in the opposite directions. They were made excentric, that is, lop-sided, by taking out a portion of the metal on one side, and replacing it either with a heavier or lighter body. The manner in which they would rotate was, therefore, known; for, not to use too scientific language, the light side moved first, and according to the relative positions of the heavy and light side when placed against the charge so the rotation took place. Thus, when the light side was resting against the bore of the gun, the rotation was exactly contrary to the direction shown in diagram No. 2; and a range of 5,566 yards was obtained from a 10-inch gun, being 916 yards farther than with a concentric shot from the same gun. The deflections to the right and left were proportionately large, according as the light side was placed to the left or right. We need not specify further; this will be sufficient to show the reason why the smooth bore with a spherical bullet never made a straight long shot, for it was not only that the bullet did not go in the direction in which it was aimed, but it did not even follow the direction in which it started. This was well shown by Mr. Robins in the experiment we commenced with. He bent the end of a gun barrel to the _left_, and aimed by the straight part. As would be naturally expected, the shot passed through the first tissue-paper screen 1½ inches to the left of the track of a bullet, which had been previously fired from a straight barrel in the same line with which the crooked barrel had been aimed, and 3 inches to the left on the second screen; but as he had predicted, and as the company could hardly have expected, on the wall which was behind, the bullet struck 14 inches to the right of the track, showing that though it had gone at first as directed by the bent portion of the barrel, yet as the bullet in being turned had _rolled against_ the right-hand side of this portion of the barrel, it had a rotatory motion impressed upon it, by which the anterior portion moved from left to right, and the bullet, after moving away from, turned back and crossed the track of the other bullet again, or was _incurvated_ to the right. We now see why spherical bullets from a smooth bore, though they may fly almost perfectly accurately a short distance, cannot be depended on in the least for a long distance, as the bullet which might strike within 1 inch at 100 yards would not strike within 2 inches at 200 yards, and still less within 3 inches at 300 yards of the mark at which it was fired.