TESTING OF PORTLAND CEMENT. 335 



substances in compression is the manner in which the 

 specimens are bedded on the compression plates. What 

 is required is an even distribution of the load over the top 

 and bottom surfaces of the block, so that the compressive 

 stress may be quite uniform. It is only when the stress 

 is thus uniform that a reliable result can be obtained, as a 

 very little extra stress on one side or the other will soon 

 cause the material to chip away on that side. 



Various methods have been employed for this purpose.. 

 Thick paper or cardboard is sometimes used, but this is 

 not an altogether satisfactory way of bedding. Metallic 

 sheet lead has also been used, but the effect is to compress 

 the lead and cause it to spread, and in doing so to drag 

 the surface of the specimen outwards and cause failure at 

 a much lower load than that due to uniform crushing alone. 



The most satisfactory way of bedding compression 

 specimens is by using plaster of paris. The compression 

 plates must first be cleaned and possibly oiled to prevent 

 the adhesion of the plaster of paris. Then the plaster is mixed 

 in sufficient quantity to the consistency of a thin paste. 

 Some of this must now be spread with a trowel on the 

 lower compression plate, so as to be not less than about 

 a quarter of an inch thick and covering rather more than 

 the area of the bottom of the specimen. The specimen is 

 now placed quickly in position on the cement and the top 

 plate screwed down tight, so as to thoroughly squeeze out any 

 superfluous plaster. After a few minutes this plaster 

 will have set sufficiently to allow the top plate to be 

 lifted. The top of the block is now smeared with plaster 

 until it is quite covered, and the top plate again screwed 

 down tight. After about five or ten minutes the block 

 will be ready for testing. 



This latter consists in simply increasing the load 

 steadily until the material fails, which it generally does 

 with comparative suddenness, little warning being given 

 by cracking of the material that failure is to be expected. 

 The failure is generally expressed as so many pounds or 

 tons per square inch or square foot of area required to 

 produce failure. 



The , shape of the fractured specimen in most cases 

 takes the form of two more or less perfect pyramids in 

 the case of cubes, and cones in the case of cylinders. 



