380 



The empirical rule which I suppose guides such engineers 

 and directors as reject the aid of chemistry, and yet profess 

 to follow some rule, and not to trust entirely to chance 

 in the selection of water, is, I suppose, to prefer such water 

 as is " soft," and to exclude " hard" springs. To a certain 

 extent this is correct, hut not without great exceptions. 

 Chloride of calcium (muriate of lime) decomposes soap, 

 and, therefore, for domestic purposes, imparts hardness in 

 such proportion that about 56 parts produce the same effect 

 as 68 of sulphate of lime and 50 of carbonate of lime. 



But since muriate of lime (chloride of calcium) is an 

 exceedingly soluble salt, it does not, as such, form any 

 portion of crust on boiling. Water containing nothing be- 

 sides muriate of lime would not " fur" at all ; and in so far 

 as the mixture of earthy salts which interferes with the use 

 of soap consists of muriate of lime (and in practice it some- 

 times forms a large portion), will the " hardness" to soap, 

 and the disposition to " fur" vary from each other. If we 

 compare by means of soap, or, what is nearly the same 

 thing, by their reputed hardness, a water containing 10 

 grains of carbonate of lime, with another containing 5 of 

 carbonate of lime, and 15 of muriate of lime, the other 

 ingredients, if any, being in each case the same, the first 

 specimen would seem only half as objectionable as the 

 second, while it is really twice as bad for the boiler. The 

 total solid matter in solution is a yet more vague method 

 of judging, because this will include not only the readily 

 soluble muriate of lime, but the salts of soda, which do 

 not form insoluble crusts at all. 



Whether we employ the terms " hard" and " soft" in the 

 popular sense, or restricted to the formation of fur, we have 

 at present no line drawn between the two, and scarcely any 

 approach to naming a standard, those superior to which 

 should be called good, those inferior, bad waters. As exact 



