384 



are to their source, whether this be a spring or surface 

 drain. I have found in such as little as three grains of 

 solid matter altogether, in the gallon, as they dissolve more 

 than they part with, in their subsequent course. Waters 

 springing from, or collected upon limestone or chalk, deposit 

 considerably from exposure to the air, agitation, and the 

 action of the sides and bottoms of brooks and of reservoirs. 

 The condition of these, therefore, is improved by being 

 conveyed in natural or other open channels, and by absorb- 

 ing in ponds and reservoirs. Many clays contain no small 

 proportion of sulphate of lime; the composition of such 

 clay as is used for puddling is, therefore, not beneath the 

 notice of the engineer. In selecting cases for analysis, it 

 is desirable either that a time of drought, when the waters 

 commonly are strongest, should be chosen, or at least ordi- 

 nary average weather. If taken during heavy rains, the 

 results of analysis will represent the water as more diluted, 

 and therefore better than it really is. Lastly, I would 

 recommend that the quantity of water should not be too 

 small. 



It is not wise to increase the difficulty of analysis, or 

 to exclude those collateral experiments on the effect of 

 exposure and subsidence, boiling, &c., which may prove as 

 useful as the analysis itself. One or two gallons is a good 

 quantity. 



Within a short period, however, I believe a few months, a 

 method has been proposed by Dr. Ritterbandt, for prevent- 

 ing incrustation even from very hard waters. His remedy 

 is muriate of ammonia, crude sal ammoniac (well known 

 to boiler-makers for a different purpose, as one ingredient 

 in a cement for iron) ; and its use in hard waters is founded 

 on the remarkable fact, that though when muriatic acid, 

 lime, carbonic acid, and ammonia, in equivalent proportions, 

 or at least excluding great excess of carbonate of ammonia, 



