CEMENT OF AN ANCIENT MOSAIC. 141 



On comparing these two analyses we find, that the cement 

 •ontains in 10,0. pacts 



Quickllnae 56'3 



Carbonic acid ....•.^.... 41 



Veg^etable or animal matter 2*7 



100-0 



Tn thia cemient, we see, the lime* if; it were employed The Kme 

 quick, ha^^resumed from the air, and in, the lapse of time, ^^^^^ **^"* 

 nearly- all the carbonic acid necest>ary for its saturation. 



This is the first time of my observing tbis fact. As I This singulai; 

 have never found the lime in mortar, however ancient, satu- 

 rated with carbonic acid, I am inclined to suppose, that the 

 vegetable or animal matter^ tbat served as a gluten, pro- How effected., 

 miotedthc absorption of carbonic acid ; or rather, that the 

 cemjCnt in question wa? made with carbonate af lime (whit- 

 ing), and not with quicklime. 



In the lattercase about 97 parts of carbonate of lime> and 

 3 of oil, glue, or cheese, must have been employed. 



In the former the e?men1; would have been composed of 

 abo^t 56; parts of quicklime to 3 of vegetable or animal 

 matter. 



It; is obvious, that these proportions, which are found at 

 present tp form the cement. in question, were not followed 

 In its preparation. 



If oil were employed, it would have increased in weight 

 in drying,; and then less than 0*03 must have been used, 

 which appears to me impossible. 



It; is more than probable therefore, that the substance era- Its probable 

 ployed was analogous to the caseous part of milk, and thea*"™^**"*'**"* 

 it would haye diminished in weight by losing, the water it 

 contained, which served to reduce to a paste the lime or 

 carbonate of lime. 



From this analysis it appears, that the cement was very 

 simple; and tbat thos^we i^ow compose on theiaame pfin* 

 •iple would become equally hard in time. 



XIY, 



