382 Bibliography. 
more fortunate in substituting pure magnesia in the place of the oxide 
of manganese ; as appears from the table, Nos. 23, 24, 25, 27, 28, 42. 
‘“‘ The positive results exhibited in the table, are more to be relied on 
and lead to more safe conclusions. The examination of them shows: 
“Ist. That silica and lime employed together, without any other ad- 
mixture, will give a hydraulic compound. Nos. 3, 8, 9, 11, 18, 34, 
35, 36. 
“© 2d. That silica, alumina and lime will produce an equally energetic 
composition. Nos. 1, 4, 6, 7, 16, 37, 38, 39. In some respects, the 
addition of alumina improves the mixture. Being insoluble in water, 
it protects the outer portions of the lime from solution, until the union 
of the latter with the silica is effected, and prevents the free permeation 
of water through the mortar, which might greatly injure its quality. 
“¢ 3d. That silica, alumina and magnesia, will form a good hydraulic 
compound, without the addition of lime, Nos. 10, 17, 18, 19. 
“ This fact may account for the existence of hydraulic qualities in 
the native carbonate of magnesia, which, like the carbonate of lime, 
may frequently contain a portion of clay in its composition. 
“4th. That silica, lime and magnesia give a better hydraulic result 
than silica and lime, Nos. 11, 12, 13, 14. This probably arises from 
the formation of double silicates, the silicic acid uniting with both lime 
and magnesia. 
“Nos. 21, 22, 41, exhibit surprising results, and lead to the inference 
which is generally deducible from the table, that the induration of 
hydraulic mortars is not to be ascribed to any one agent, nor ever to 
precisely the same causes ; though, in most cases, it is owing to the 
formation of a silicate. 
“¢ With the view of discovering (says Lt. Wright) the constituents of 
some of our hydraulic limes, I submitted for‘analysis to Dr. Jackson of 
Boston, two specimens of limestone, obtained from different localities 
and possessed of different properties. ‘They both furnish after calcina- 
tion, products, which it is necessary to pulverize, in order to prepare 
them for use, and are both known in commerce as hydraulic cements ; 
one called Barnes’ Connecticut cement, from Southington, Conn., the. 
other Lawrence’s Rosendale cement, from the vicinity of Rondout, 
New York. 
“The second is esteemed the best cement for use, setting in five 
minutes after immersion in water. 
‘“‘ The results of the two analyses were as follows :—One hundred 
grains of the Connecticut cement, being thoroughly dried at 212° F. 
yielded on chemical analysis— 
sgl 
