EFFECT OF SURFACE CONDENSATION

184

“No other alternative was therefore left, noris there at present any other as far as Mr. Jackhas been able to learn, but to feed the boilerswith a portion of salt water sufficient to keepa thin incrustation over the surface of the iron.It was suggested that the deposit was nothingelse than rust or oxide of iron, and that it wasformed by the chlorine present in the smallproportion of salt water, which would combinewith the iron to form chloride of iron ; and thisbeing readily decomposed by oxygen, oxide ofiron would result. The difficulty here, however,was to know whence the oxygen was obtained;for the quantity of air entering with the feedwater must have been very small indeed. Itwas also suggested that hydrochloric acid mightbe present, from the small quantity of sea waterthat may have found its way into the boilers;hut then the difficulty was to know where aquantity of the acid was to come from, sufficientto act over such an extended surface, and asrapidly as the results showed.

“ It was found however by Mr. Hollo, one ofMr. Jack s partners, that in a pair of boilers ata sugar refinery there was the same brownishdeposit adhering all over the boilers, and thoseparts subjected to the greatest heat were‘ pitted in precisely the same manner as thesecond lot of marine boilers previously de-signated by the letter B. Exactly the sameeffects were being produced. These boilerswere supplied with the same water over andover again, a small quantity of fresh waterbeing added from time to time to make up forthe loss. As the steam was passed only throughiron pipes for melting the sugar, the damage tothe boilers could not result either from thesteam coming in contact with a soft metal, or

from any lubricating material. The boilerswere of the Cornish construction with one flue,and were worked at about the same pressureas the marine boilers B already referred to, say20 lbs. per square inch pressure. A pair ofboilers of exactly the same construction, placedalongside the first pair, and working at aboutthe same pressure, but fed with water whichhad not been distilled, were then examined, tolearn what state they were in. But althoughput in about the same time as the two firstexamined, these boilers were found in good con-dition and likely to last for years, as there wasnot a sign of corrosion or ‘ pittingwhereasthe two boilers working with distilled waterhad to be repaired.

“ The practical knowledge thus acquirednecessarily led to the conclusion that the dis-tilled water itself was the cause of the corrosion,instead of any galvanic action or any fattyacid. In reference to the question whetherdistilled water has any particular action onmetals, the chemist Berthier found that nodularprotuberances deposited on iron pipes contain-ing distilled water consisted of 21 per cent, ofprotoxide of iron, 58 per cent, of peroxide ofiron, 5 per cent, of carbonic acid, 14| per cent,of water, and 1J- per cent, of silica. The ironpipes contained also a pulverulent substance,which could be produced at pleasure with dis-tilled water to which a trace of carbonate ofsoda and common salt had been added, but notwith an addition of caustic alkali. Distilledwater is known to act powerfully on lead, andthis action is attributed by Dr. Clark to theremarkable property that distilled water has,as compared with ordinary water, of, dis-solving free carbonic acid.