THE EXCAVATIONS 61 



curved tapering imbrices. In tlie tegulae nailholes are 

 found wliicli seem to show that nails of oblong section 

 were used, and an abundance of iron nails has been found on 

 the site. Some of the bricks measure 10| in. by lOfin. 

 by2|in. 



Under one of the large blocks at the west gate an ex- 

 cellent specimen of the mortar (still white and hard, 

 though deteriorating) may be seen in situ. I submitted 

 a specimen to Professor Boyd Dawkins, and he pronounces 

 it to be made with sand from the millstone grit of the 

 neighbourhood. '■I Mr. Francis Jones has made an analysis 

 of this mortar. The analysis gives the following results : — 



Silica 85-47 



Lime (CaO) 5-08 



Iron and Alumina (Fe^iOg and ALOa) 266 



Carbon dioxide 282 



Water (dried at 200°C.) 104 



Magnesia (MgO) Trace. 



Alkalies, etc. (not det.) 2'93 



10000 

 There was more lime than corresponded to the amount 

 of carbon dioxide found, but as sulphuric a^id is also 

 present, the remaining lime is no doubt present as sul- 

 phate and also as silicate."^ 



It is interesting to remember, in this connection, that 



71. Vitruvius devoted a whole chapter to the question of the selection 

 of sand. De Arch., ii., 4. 



72. As affording an interesting case for comparison I give the figures 

 of the analysis of the mortar found in the walls of Hadrian's villa. 

 They are as follows : — Silica 4110, Alumina 1470, Lime 15-50. Ferric 

 oxide 492, Magnesia OSO, Carbon dioxide ll'SO. Potash 101, Soda 212, 

 Organic matter 228, Water 520, Total 98-73. ' (See W. Wallace : On 

 ancient mortars, Chem. News, 1865, vol. xi., p. 185, and Dingier' s 

 Poly tech. Jrnl., 1865, vol. clxxviii., p. 372. See also Thorpe, Diet. 

 Appl. Chem., vol. i., p. 467.) The cement of the mosaic on the Baths 

 of Caracalla at Rome contains 25- 19 per cent, of lime. Mortar from 

 the Pnyx at Athens has 4570 per cent, of lime. It is not easy to say if 

 any of the original lime has been washed away from the specimen of 

 Melandra mortar analysed by !Mr. Jones. 



