386 CEMENTS, LIMES, AND PLASTERS. 



some consideration. 'Low lime will invariably mean low-testing cements, 

 and in the present state of the industry, low-testing cements are not 

 easily marketed. A low-lime content is also the cause, in part, of the 

 "dusting" of clinker in the vertical kiln. Le Chatelier found that the 

 dicalcic silicate (2CaO,SiC>2) possesses the property of spontaneously 

 disintegrating on cooling. If the lime content of the mixture be carried 

 too low, therefore, the clinker will fall to dust in the kiln, owing to the 

 production of this unstable dicalcic silicate. 



Magnesia. The question as to the percentage of magnesia allow- 

 able in a Portland cement has given rise to serious controversy for 

 many years. In Europe the tendency has been to keep it below 3 per 

 cent; but in this country, largely because of the results attained by 

 Lehigh Valley cements above this limit, 4 or 5 per cent has been con- 

 sidered the allowable maximum. All this discussion .was carried on 

 under the idea that magnesia was either inert or positively harmful in a 

 Portland cement. 



Recent experiments by Prof. Newberry, however, have proven that 

 an entirely satisfactory cement can be made carrying as high as 10 

 per cent of magnesia, if due care be given to the mixing and burning. 

 This might have been expected, both on theoretical grounds and be- 

 cause of the evidently active nature of magnesia in even the highest- 

 burned natural cements, as pointed out on pages 198-200. At present 

 it seems safe to say that magnesia can be considered equivalent to lime 

 in its action, if due allowance be made for the difference in their com- 

 bining weights. It is therefore theoretically possible to prepare a series 

 of lime-magnesia Portlands, parallel to our present lime Portlands; 

 and it is probable enough that in a few years some move will be made 

 in this direction. But it must be borne in mind that a lime-magnesia 

 Portland will probably differ in important respects from our present lime 

 Portlands, and that it will therefore be inadvisable to group the two 

 types of cement under the same general name. For this reason, in 

 the present volume, the term Portland has been restricted by defini- 

 tion to apply only to cements carrying less than 5 per cent of magnesia 

 (MgO). 



Silica. It is commonly considered that the ultimate strength of 

 the cement depends in large part upon the amount of calcium trisilicate 

 it contains. Within certain limits, therefore, any increase in the per- 

 centage of silica in the mixture will increase the strength of the cement. 

 On the other hand, an increase in silica will usually imply a decrease 

 in alumina and iron oxide, and this in turn will cause the cement to 

 be slow-setting (which is an advantage), but hard to clinker. 



