1852.] On Filtering the Waters of Tanks. 475 



upwards to C. If these spaces be ten feet deep only, this gives twenty 

 feet of filtering distance, in every inch of which the water will leave 

 some of its impurities. A small wall and sluice outside of A. would 

 cut off the communication and enable us to clear out our filtering 

 apoaratus and re-fill it with fresh materials. It is probable that no 

 water would require more filtering than this to come perfectly limpid 

 into the reservoir, but if any did so, another pair of walls might be 

 added. They may be tolerably close ; say just far enough apart to 

 allow a man to work in clearing out the material when it requires 

 changing, for it is to be noted that the efficiency of this filter depends 

 upon its depth, and not on its breadth at all. 



It is evident that walls may be built to any extent required, either 

 merely to inclose a ghat, or a corner of a tank, or across a whole side 

 of it ; and that arrangements may easily be made for preventing the 

 fouling of the limpid water, when filtered, by those who take it for use. 

 A modification of this which might be adopted in private tanks or 

 even in public ones, would be to have sheet iron caissoons made, one 

 within the other, and placed in a tank, so that the central space 

 should always be a well of limpid water. 



But as above remarked, clear water is not always pure water, and 

 that of the river for instance, though filtered till perfectly pellucid, 

 would no doubt still contain animal, vegetable and saline matters 

 which being held in solution must be separated by some natural che- 

 mical process, and this may be called chemical filtering. We do not 

 know what the impurities of our water are, but we will set them down 

 generally as animal or vegetable matters and saline ingredients, such 

 as phosphates, carbonates, sulphates and muriates of various bases. 



We have then first to determine by varied experiments which of the 

 cheap and easily obtained substances in our second column is likely 

 best to answer our purpose, and there is no doubt but that some of 

 them simply, as the iron stones in various forms ; or mixtures of them, 

 as chalk with kunkur gravel, or chalk with the coarsely pulverised 

 basalts ; or with the black sand and the like, will decompose the saline 

 matters, and at the same time, frequently cause also the separation 

 of the animal or vegetable matter, or of the greatest part of it. A 

 familiar instance of the use of iron earths is well known to chemists 

 in the use of the water of the Seine at Paris, which is what is called 



