BULLETIN OF THE UNITED STATES FI8H COMMISSION. 151 



32.-THE BOWER-BARFF RUSTLESS IRON. 



By J. II. KIDDER, in. D. 



[From letters to Prof. S. F. Baird.] 



I return herewith the specimens of iron pipe treated by the " Bower- 

 Barff rustless iron" process, with memorandum of the results of exami- 

 nation and the following note upon the process itself: 



Dr. Percy first pointed out the fact that Russian sheet-iron is much 

 less affected by rust than English, because it has been accidentally 

 coated with magnetic oxide (Fe 3 4 ). 



Professor Barff, prior to the year 1876, first produced a coating of 

 magnetic oxide upon iron for the purpose of preventing rust. lie ex- 

 posed the heated metal to superheated steam, which at high tempera- 

 tares is decomposed, giving up its oxygen to the iron, while hydrogen 

 escapes. 



Mr. George Bower improved upon Barff s method (about the year 1880) 

 by passing the products of imperfect combustion (carbonic oxide, &c.j 

 through highly heated chambers containing air in slight excess of the 

 quantity required to convert carbonic oxide into carbonic acid. The 

 resulting carbonic acid, heated by combustion, enters a chamber con- 

 taining the articles to be coated, raising them to a high temperature. 

 The iron takes oxygen from carbonic acid and becomes magnetic oxide 

 at its surface, covered by a film of sesquioxide, while carbonic oxide 

 escapes. This process, called the "oxidizing process," is continued for 

 half an hour, when the air inlet is closed and only carbonic oxide ad- 

 mitted to the iron. Carbonic oxide (CO) takes oxygen from the film of 

 sesquioxide covering the magnetic oxide, reducing all of the coating to 

 magnetic oxide. This is called the "deoxidizing process," and lusts 

 twenty minutes. The two may be repeated according to the thickness 

 of oxide desired. 



There is no reasonable doubt that magnetic oxide of iron is unaf- 

 fected by exposure to air or fresh or salt water, or that, if thoroughly 

 and completely applied, a coating of this oxide will protect iron surfaces 

 from rust. The specimens experimented upon appear not to have been 

 thoroughly coated, especially in the screw-threads, and there is some 

 reason for suspecting galvanic action upon the specimen A (in salt 

 water). 1 believe that the authorities at the Portsmouth dock yard 

 have already decided that such action will occur when coated and nn- 

 coated specimens are exposed together to sea-water. 



For further experiment in this direction I recommend that some strips 

 of iron be furnished which have been completely covered with the mag- 

 netic oxide, exposing no uuoxidized surfaces. The strips may con- 

 veniently l»e 3 inches wide by 4 inches long, but should not be sheared 



