LIME AND CEMENT INDUSTRIES V31 



9 



to the long arm. The calipers are readily moved in any direc- 

 tion, and the micrometer is read in the usual manner. One revo- 

 lution of the screw equals .5 mm ( -^\ of an inch), and readings 

 on the head are made at ^o mm (-jtjW of an inch). The speci- 

 men is placed on a small platform, between the lever and the 

 screw. The points of the calipers are set on center marks drilled 

 into small glass plates let into the specimens. 



The width between the caliper points is made equal to 95 mm 

 (3f inches) in each of these instruments, thus very much simpli- 

 fying the computation for length. For instance, if the screw 

 reads 9.56 revolutions, the absolute length of the specimens is 

 9.56 -=- 2 + 95 mm = 99.78 mm. The specimens are made in 

 small metal frames, just as the standard specimens for tension. 

 It is necessary, however, to turn the molds over repeatedly, and 

 treat both the upper and under surfaces alike. If this is not 

 done, and the upper surface becomes rather thick and smooth, 

 which a repeated striking off with the trowel will accomplish, it 

 may easily happen that the lower layers remain loose and porous, 

 causing a distortion of the specimen, which may lead to consid- 

 erable errors. The positions for the center-mark plates are pro- 

 vided for in the forms, and these plates may, therefore, be ce- 

 mented into place as soon as the specimens are removed from 

 the molds. To measure a specimen requires but a few minutes, 

 the apparatus being very easy to manipulate.. 



In the following table, F and G are two cements which were 

 tested for tensile strength in a 1-3 mortar, and showed but small 

 strength. It will be noted that these two inferior brands showed 

 an extraordinary degree of shrinkage, making them unfit for 

 decorative purposes and laying of face stones. This extraordi- 

 nary shrinkage explains the cracks shown on so many ornamental 

 surfaces, artificial stones and plates, which always have either a 

 neat cement or a mixture low in sand at their surface. The 

 preference for a really good brand of cement for this purpose is 

 thus explained. The table furthermore shows that the commonly 

 adopted theory regarding a uniform relation between expansion 

 when hardening under water and shrinkage when hardening in 

 air is erroneous. 



