SOME NOTES ON ANCIENT MORTAES. 39 



mortar. This is formed, according to Petzold, by the action 

 of slaked lime on quartz sand, while according to Winkler it 

 is dae to the presence of alumina in the quicklime used for 

 the mortar, and is formed when the limestone is burnt. 

 Feichlinger, however, considers it quite possible that the 

 formation of calcium silicate may take place at the point of 

 contact between the mortar and the building stone. 



In judging of a mortar, the relation of lime to sand is of 

 importance. This relation varies according to the quality 

 of the sand, for fine sand required less lime than coarse. In 

 the mortar under consideration the sand was unusually 

 coarse, and large pieces of quartz could easily be recognised. 

 It has been estimated that a good lime-mortar should con- 

 tain 15 per cent, of slaked lime in the dried mass. The 

 amount of hydrated lime in the above is only a little over 

 7 per cent. The above remarks show that this ancient 

 Roman mortar had very inferior properties. This is con- 

 firmed by the fact that it easily crumbles away. The per- 

 centage of alumina and ferric oxide appears to be high, but 

 the quantity of magnesia is small. 



I venture to bring before you the composition of some 

 mortars used by the Romans in this country, and as an 

 example I will take the mortar from beneath some tesse- 

 lated pavement which was unearthed while excavating for 

 the foundations of the new Guildhall at Grloucester during 

 the summer of 1890. This mortar was fairly hard and 

 thoroughly hydraulic, and when examined under the micro- 

 scope was found to consist of a white, or rather a pink 

 coloured matrix cementing together fragments of burnt clay 

 (crushed bricks, tiles, or pottery) or pozzuolana of a bright 

 red or reddish-brown colour and rounded grains of pure, 

 colourless quartz sand very much resembling the standard 

 cement testing sand from Leighton Buzzard, or some coarse 



