EECEMENTATION. 



13 



have no particles of a colloid nature, and therefore no binding power, 

 soaking or kneading will not develop the slightest increase. In carry- 

 ing out the cementation test the doughs are now invariably aged for 

 at least twenty-four hours, and the workman is trained to knead the 

 dough always in the same manner for about the same length of time. 

 Careful attention to these details has resulted in a great improvement 

 in the value of the test, both in a lessened percentage of variation in 

 different test pieces of the same series, and also in the agreement of 

 the tests with the results of service. The percentage of error from 

 variation is about the same as in cement testing if carried out with the 

 same care. 



Several criticisms have been made upon the cementation test, the 

 opinion offered being to the effect that the results are too uncertain to 

 be of value. The persons making these criticisms had not, however, 



10 2.0 X> 





SO 60 7O 00 90 100 HO ISO ISO MO ISO I6O /TO ISO /9Q 200 

 AGE OF DOUGH IN HOURS. 



FIG. 6. Diagram showing prolonged effect of water on cementing value. 



availed themselves of these later improvements in the manner of mak- 

 ing the test pieces. On the other hand, it can not be too strongly 

 urged that the object of the test is to gain information which will be 

 of use to road builders and not merely to accumulate data of abstract 

 interest. Rocks are far from being homogeneous materials, and in 

 addition the cementing power itself is such a variable quantity that 

 it may be said that the test has served its purpose if it does no more 

 than distinguish high binding material from that which is only fair, 

 and this in turn from that which has little or no binding power at all. 

 We come now to the bearing of the test upon the results of actual 

 service. There have been few instances in the experience of the 

 laboratory where the results actually attained upon the road have not 

 agreed with the results obtained in the laboratory; the few exceptions 

 to this rule occurred before the investigations were completed, the 

 results of which are given in Table II. In many of the Southern and 



