OF WATER FOR ORGANIC MATTER. 61 



much organic matter, it is often very difficult to know 

 when to stop, as the brown masks the red colour. But 

 this must not bring discouragement, as experience will 

 teach exactness ; and if it does not allow exactness, it then 

 will be no great loss in such cases, in a sanitary point of 

 view, as in them there is decidedly too much organic 

 matter, and the water may be condemned. From the 

 same point of view, it is of less consequence whether the 

 amount be a minute quantity more or less. It is well to 

 make broad lines of distinction, and to condemn freely 

 when there is rational hope of obtaining a purer water. 



^' The amount of available oxygen in one of the solutions 

 of chameleon described is 0'0005, which may be either 

 grammes or grains, or almost exactly one of oxygen in 

 2000 of the solution. If we wish to know the amount of 

 oxygen used, we simply divide the amount of chameleon 

 by 2000. It is probable that the use of the oxygen column 

 is the only exact method of recording the results ; and it 

 is probable that the amount of oxygen required is the 

 only exact measure of the impurity of the substances. We 

 obtain in this number the amount of oxygen which is 

 required for purification. 



" Whilst looking over this plan in manuscript, I received 

 Dr. Miller^s paper, proposing also to use oxygen; he 

 prefers a solution of chameleon containing 3*95 grains, 

 equal to 1 grain of available oxygen to 10,000 of water, 

 or 1 cub. cent. = 0*001 gramme, or say 1 milligramme of 

 oxygen. This is a very convenient method; but it has 

 been thought proper to keep the Tables as given, for various 

 reasons, chiefly arising from this, that it has not been 

 found so convenient to keep or to titre a very weak solu- 

 tion. In some cases a very strong solution is required ; 

 we can easily dilute it, but we cannot concentrate it 

 readily. After estimating its stength, we dilute it (if we 

 require a weak solution) to any amount we think proper ; 



