150 Scientific Proceedings, Royal Dublin Society. 
if the water is employed in a boiler to which air has frequent 
access, it would, of course, be impossible to assign any limit to the 
oxidation that may take place. The corrosion produced by the 
dissolved gases alone might be a long time in operation without 
doing much harm, if it were evenly distributed ; this, however, is 
not the case, the corrosive action being for the most part confined 
to a few spots, and thus serious injury often results from a com- 
paratively small amount of oxidation. The most convenient re- 
medy is one that has been for many years employed in soft water 
districts—viz., the addition of a small quantity of sodium or 
potassium carbonate to the water. I have found that the smallest 
proportion of the alkaline salt that can be employed with success 
is one part by weight to two hundred of water. A piece of bright 
iron partly immersed two years ago in water, containing 0°5 per 
cent. of sodium carbonate, is still quite bright, and even that part 
of the iron which was not immersed, but freely exposed to the 
air, is perfectly free from rust. 
—— 
