Tarr — Origin of the Chert ia the Burlington Limestone. 435 



sea water were added, Mg being the only bivalent ion used, 

 but if a solution consisting mostly or entirely of univalent ions 

 could cause coagulation then one containing bivalent ions 

 would be still more effective. One of the solutions used con- 

 tained NaCl, K^SO^, and Mjj^SO^, and the other one contained 

 only NaCl. 



Preliminary experiments with dilute solutions of salt and 

 silicic acid of unknown strength showed that the silicic acid 

 would be coagulated at once. Then solutions like those 

 described above were used iu conjunction with a dilute solu- 

 tion of water glass, hydrous sodium silicate, the amount of 

 silica added being from 13 to 27 parts per 1,000,000. This 

 represents about the range of the silica content of a great many 

 of the present rivers. Many rivers in the United States are 

 carrying more than 30 parts per 1,000,000 of silica. Some of 

 these are draining areas where the rate of mechanical erosion 

 is low. That the method of precipitation is effective for a low^ 

 silica content is shown by experiment three. 



Experiment 1. — A solution representing sea water was made, 

 containing 34,400 parts of solid matter per million, which is 

 the average strength of sea water. The salts used in making 

 the solution were NaCl, MgSO^, and K^SO^, the relative 

 amounts of each taken being those found in actual sea w^ater. 

 Silica in the form of water glass was then added to the solu- 

 tion. The amount of silica added was at the rate of 27*07 

 parts per million. The result of the experiment was an immedi- 

 ate and heavy precipitate of gelatinous silica. jN^o subsequent 

 change occurred in it. 



Experiment £. — In this experiment the solution of sea water 

 used was prepared by adding only NaCl, but at the normal 

 strength of 34,400 parts per million. Then the same amount 

 of silica was added as in the first experiment, i. e., at the rate 

 of 27'07 parts per million. The result was an immediate pre- 

 cipitation of silica gel, but a much less heavy one than in the 

 first experiment. At the end of two hours it had coagulated 

 into small globules on the bottom of the container, but after 

 that no further change took place. 



Experiment 3. — In the third experiment sea water as 

 described in experiment 1 was again used and a smallei* 

 amount of silica was added. The amount added was at the 

 rate of 13*53 parts per million or just half the amount that was 

 used in the first experiment. An immediate and heavy pre- 

 cipitate occurred. So change occurred in it upon standing. 



Experiment J^. — In experiment 4, experiment 2 was repeated 

 except that twice the amount of silica was used, i. e., 54*14 

 parts per million. The precipitate was immediate but not 

 heavy. No subsequent change took place upon standing. 



