116 BACTERIA AND THE DISINTEGRATION OF CEMENT, 



cement would, on setting, bring about the formation of calcium 

 silicate instead of free lime. Barth, by facing the disintegrating 

 canals with a cement containing a higher percentage of silica, 

 showed the importance of minimising the free lime in hardened 

 cement subjected to the action of water. Schiffnerf concluded 

 from the results of experiments conducted upon the Bonn reser- 

 voir (the disintegrated cement of which Stutzer investigated) 

 that no calcareous material is capable of permanently resisting 

 the action of running water. " As regards protective coatings 

 for cement-lined reservoirs, an experience of 33 months teaches 

 that the fluorine preparations sold for this purpose give satisfac- 

 tory results; and oxalic acid and ammonium oxalate (the latter 

 being best) also considerably retard the corrosion. Similarly 

 asphaltum varnish preserves the cement." 



Postscript {added May 16th, 1901). — In the discussion that 

 followed the reading of this paper, it was suggested that the 

 lime had in the experiment been dissolved as a silicate, because 

 it seemed incredible that there could be free lime in the exhibited 

 sample of disintegrated cement. I was led to the conclusion that 

 it was there as free lime b}'- the relatively abundant formation of 

 calcium carbonate crystals on the walls of the culture flasks. To 

 test the matter thoroughly, however, I repeated the experiment, 

 using 5 grms. of sample and 1 litre of boiled and cooled neutral 

 distilled water. After three days the solution was filtered, 

 neutralised with standard acid and evaporated to dryness. The 

 silica was coagulated by repeated evaporation with concentrated 

 hydrochloric acid, followed by heating at 130° C. The results 

 showed that there had been dissolved in the water 39 mgrm. 

 calcium oxide and 1 mgrm. silica, a ratio of CaO : SiOo : : 42 : 1. 

 This is enough to confirm the supposition that the lime is dis- 

 solved in the free state and not as silicate. On calculating the 

 calcium oxide to percentage in the sample, 0*78% is obtained. I 



t Schiffner, Thonind. Zeit. 24 (100, Suppl.) 1320-i;^21; Abstract in Jour. 

 Soc. Chem. Industry, xix. 1114. 



