The abrasion test may also be made by putting the weighed bricks on a grinding table covered with sand and water, and noting the weight before and after grinding. This last method seems to us to be decidedly the best, provided the bricks be weighted, that the weight is constant, that the feed of sand and water is uniform, and that the bricks to be tested are placed equidistant from the centre of the turning table. If this last point be not attended to, it will be obvious that in course of the revolutions the sand will tend to accumulate towards the centre of the table, and the bricks placed in that vicinity would receive more than their fair share of abrasion, as compared with those bricks situated near the edge of the table. Conversely, those bricks near the periphery would be subjected to greater grinding action, from the circumstance that the table would move faster underneath them than under those bricks nearer the centre of the table. The bricks should certainly be weighted in such abrasion tests, and it seems desirable that the weights should be so adjusted that the weight of the brick is also taken into account. It is obvious that the abrading action of, say, street traffic, will be the same on a brick, no matter what the latter weighs, depending on the area of surface exposed to traffic. And if we experiment with one brick, weighing say 7 lbs., and another weighing 14 lbs., the greater weight of the latter, will (_cæteris paribus_), by the abrasion tests as usually adopted, give a much higher result than would the lighter brick. On the other hand, if the 7 lbs. brick be weighted another 7 lbs., then the results would be strictly comparable, provided always that the area exposed to abrasion in each case be the same, and that the other conditions we have laid down are strictly observed. Knowing as we do that the rough and ready method of “rattling” cannot possibly give truly comparative results, we do not intend to enlarge much on the results of the American tests; but the following are suggestive as shewing the connexion between the tests for absorption, rattling, and strength combined. Some valuable and interesting tests were recently made by the Ohio Geological Survey, to determine the relative merits of fire-clays and shales for the manufacture of paving bricks, as well as the influence, if any, of the method of manufacture adopted. Twenty-two varieties of shale bricks, or bricks the largest constituent of which is shale, were grouped together: fifteen varieties of fire-clay brick; four varieties composed of shale and fire-clay mixed in equal proportion; and three varieties made from Ohio River sedimentary clays. The averages of these four classes of results were as follow:--