Dean — Formation of Missouri Cherts. 413 



dolomite, an acid carbonate is formed which like all salts 

 is more or less readily dissociated as shown : 



CaCO, + H,0 + CO, ^z± Ca (HCO s ) B 

 Ca (HCO s ), ^z± Ca++ + 2HC03 



The positive calcium ions so formed are very efficacious 

 in precipitating colloidal silica, even more so than the 

 small amount of hydrogen ions formed from the disso- 

 ciation of carbonic acid, since the precipitating power 

 increases with the valence. A colloidal silica solution 

 prepared from water-glass and hydrochloric acid was 

 purified by dialysis until free from chlorides. To test 

 the effect of calcium carbonate on silica hydrosols and 

 determine if the reaction was due to the presence of car- 

 bon dioxide, two samples of colloidal silica solution were 

 taken, one was freed from carbon dioxide and the other 

 saturated with it. To both solutions was added ground 

 calcite with the result that the solution free from carbon 

 dioxide remained stable for over a year while the one con- 

 taining carbon dioxide had precipitated within an hour. 

 This shows conclusively that silica hydrosols will be pre- 

 cipitated by calcium carbonate in the presence of carbon 

 dioxide while in the absence of carbon dioxide the calcite 

 acts as a remarkable stabilizer, since the original silica 

 sol precipitated within a few weeks. Several ordinary 

 dolomites and calcareous shales were tried in the same 

 way both in powdered form and in coarse fragments, all 

 of which produced the same results as the pure calcium 

 carbonate. Doubtless if the solution were made very 

 dilute and a long period of time allowed small pieces of 

 limestone could be silicified as the silica is precipitated 

 where the calcium carbonate goes into solution. This is 

 probably the explanation of Church's experiment in which 

 a coral was silicified. We have then an adequate 

 explanation of the replacement of limestone as well as 

 the precipitation of silica. It is of course true that this 

 process would tend to localize the cherts along bedding 

 planes and channels for the circulation of ground water, 

 and they are found in this connection, but they are also 

 found in the shale layers and in the solid limestone away 

 from any fissures. Their occurrence in the shale bands 

 is probably due to the fact that these shales are the local 

 source of the silica. The occurrence of the nodules in 



