It should be understood that the brick tested were much more nearly, uniform in quality than would be obtained in taking brick at random from piles along the street, since the selection was made with a view to securing uniformity. The variation between duplicate rattler tests is therefore smaller than may be expected in tests of brick from the street, and the uniformity in the modulus is also greater. There was greater freedom from the accidental variations which frequently affect the rattler test. Although the rattler test was fairly. discriminating in determining quality, the results must not be taken to indicate that the rattler test may be used to settle the exact order of various makes of brick with respect to wearing quality; it should rather be considered a means of determining whether a brick is up to a required standard. The objection sometimes made that the rattler test does not easily permit determination of variation in individual bricks was not considered in the investigation, since so careful a selection of brick was made. The information given in the cross-breaking and absorption tests is valuable, and the usefulness of these tests is shown, particularly in connection with the study of the qualities of different grades of the same make of brick.
The effect of size of the brick upon the loss found in the rattler was not included in these tests. It is established that the brick size will sustain a greater loss than the block size of the' same grade and qual ity. This excess is due to the greater relative exposure of the corners which chip off more or less, and to the greater proportional wearing surface exposed in the brick size. The amount of this difference de pends upon various conditions, but with good material the brick size may be expected to lose, say, 3 per cent more than the block size. Of course, only a part of this difference would show up in the wear of pavements constructed with brick of the two sizes. The effect of accidental differ ences, or of variations in the quality of the shot, or of the smoothness or other conditions of the rattler was not studied, and will not be dis cussed here.
A study of Table II shows that the best grade of brick received in the first 450 revolutions of the rattler test from 47 to 53 per cent of the total loss and that the poorer grades lost during this stage a smaller per centage of their total loss, as little as 30 per cent in some cases. Simil arly at 900 revolutions, the better grades had received 67 to 77 per cent of their total loss, while the poorer grades had received a smaller propor tion of their total loss. It seems that the better grades wear more slowly, comparatively, after the corners are rounded off ; and the poorer grades continue to grind off or break up during the latter part of the test.
This extensive series of tests gives data on a wide range of brick and enables comparison to be made of a wide variety of conditions. It is val uable for making a study of the properties of paving brick, as well as for making a comparison of the various tests and requirements for pav ing brick. It will be seen from Fig. 3 that the best grades of brick in the first ten makes of brick, as shown by the samples tested, are of excellent quality and will make a durable and satisfactory pavement. The remaining makes are less valuable as paving material, and be sides many of these may not be judged from their general characteristics since they vary widely with slight changes in general appearance. The rattler test is a fairly satisfactory test for a particular make of brick in picking out the best degree of burning, etc., but in determining the ranking of several makes of brick it should be supplemented with the transverse and crushing tests. The absorption test is of value in study ing the characteristics of a given make of brick and in judging of the effect of changes in the amount of burning.
Reference may well be made to the information which a careful ob server will obtain in such a series of tests by means of the ocular exam ination of the structure and appearance of the brick. It suggests the desirability of a study by inspection of the structure and behavior of the brick in connection with the tests made on the brick to be used.