212 REIBLING AND REYES. 



and not to the method employed in testing. We have found that the 

 testing of properly seasoned cements of the best manufacture does not 

 necessitate such variations. With good cements, a plastic paste of nor- 

 mal consistency is readily obtained with from 19 to 23 per cent of 

 water, and competent testers readily secure concordant results, serious 

 discrepancies occui'ing only when outside influences had brought about 

 changes in the setting properties of the cement before the tests were 

 made. 



Portland cements which requires 28 per cent of water, or which can 

 not readily be gauged into a plastic paste of normal consistency without 

 flooding them with water, deserve no further consideration as they will 

 also lack in sand-carrying capacity and troweling properties and, there- 

 fore, will be but poorly adapted to meet the modem requirements of 

 reinforced concrete. 



It is also evident that a cement, the development of the setting prop- 

 erties of which can not be controlled within reasonable limits by the 

 carefull}^ regulated conditions imposed by this test, can not be guar- 

 anteed to give satisfactory service under the variable conditions met 

 with when it is to be used for construction work. 



It also has been stated that no standard method for determining the 

 rate of set really measures the exact time of the beginning or ending 

 of this early process of hardening. This is true, but fortunately the 

 setting properties of Portland cement become a consideration of par- 

 amount importance only when the set is abnormally quick or abnor- 

 mally slow. In quick-setting cements the process when once it begins 

 proceeds to completion very rapidly. The cement can harden but very 

 little when thB Yicat needle method is employed, before the set needle 

 ceases to penetrate the required distance. Therefore quick, normal and . 

 extremely slow-setting cements are readily identified by this method,- and 

 for all practical purposes no more accurate determination is necessary. 



However, it sometime happens that the cement paste held in the 

 conical, hard-rubber ring of the apparatus fails to harden equally 

 throughout its mass. It also is impossible to eliminate air bubbles 

 from the neat mortar. Therefore, in our research work when the 

 greatest accuracy obtainable was desired, we took the average of three 

 penetrations of the set needle which was applied at points equidistant 

 from the center and the circumference of the upper surface of the 

 cement so as to form the points of an equilateral triangle. 



