CHEMICAL AND PHYSICAL PAPERS. 91 



machinery and dies and the mixture of sand with the clay to reduce 

 its richness finally overcame the difficulty of structure and met the 

 specified requirement as to transverse strength. 



Transverse tests were made by placing the brick edgewise on knife 

 edges, six inches apart, with the load applied at the center of top edge 

 through a third knife edge. Crushing tests were made on half brick 

 broken in the transverse tests, these being bedded edgewise in plaster 

 of Paris on both the top and bottom. The specimen was placed in 

 the machine, an Olsen of 100,000 pounds capacity, before the plaster 

 was set ; then under a light load of one to two thousand pounds it was 

 allowed to rest for about ten minutes ; then power was applied slowly 

 till failure resulted. A few brick have been broken with strawboard 

 or with soft pine cushions. Though more rapid, preference is given 

 to the plaster bedding. 



The rattler test is designed to furnish evidence of the ability of 

 brick to withstand impact and abrasion — the blows from horses' feet 

 that will chip off edges and corners and the wear from both horses 

 and vehicles. The standard rattler is a cast-iron barrel with fourteen- 

 sided polygons for ends and fourteen staves, set with small cracks be- 

 tween them. It is twenty inches long and twenty-eight inches in 

 diameter, and is rotated with its axis horizontal at from twenty-seven 

 to thirty-two revolutions per minute. Under the old method, enough 

 brick to equal fifteen per cent, of the volume of the rattler, somewhere 

 from twenty-two to twenty-six brick of common sizes, constituted the 

 sole charge, and these were tumbled around for a maximum of 1800 

 revolutions. 



Table III and figure 1 give the results of tests made by this method. 

 Behind the Lawrence records there also is a story of many months' 

 experimentation in modifying machinery, in mixing materials for dif- 

 ferent parts of the clay-bank, and in regulating the process of drying 

 and burning. No. 39 was the first brick to come within the require- 

 ments set by the specifications for the first street paving in Lawrence. 

 Tests of later brick are better still. Every manufacturer of paving 

 brick has his own problem in adapting his methods to his materials. 

 Clay-banks are not alike, and what will work in one place will not in 

 another. Moreover, as the character of a shale-bed or clay-bank may 

 change in a few feet of distance, the brickmaker must keep vigil if he 

 is to maintain a high and uniform standard. 



A rattler test, when platted as in figure 5, is significant in several 

 ways. A curve that rises rapidly during the first 200 or 400 revolu- 

 tions indicates a large loss due to the chij^ping off of edges and cor- 

 ners. This means brittleness. A flat curve here means relative greater 

 toughness. If a curve continues to rise rapidly and the total loss is 



