MECHANICS AND USEFUL ARTS. 35 



not greater than from 10 to 18 tuns per square inch, notwithstanding 

 that the average strength may be 22 tuns, and the maximum 27 tuns. 

 And the loss of this strength in punching the rivet holes is not mere- 

 ly that of the iron cut out, but the punch is found to sensibly injure 

 that which remains. Mr. Fairbairn's well-known and frequently 

 verified ratio of 56 to 100, as the strength of a single riveted joint to 

 that of the unpunched plate, must be always admitted in calculations 

 of the strength of riveted boilers. The 40-horse Lancashire boiler, 

 seven feet in diameter, will thus be often found to have an ultimate 

 strength not greater, when new, than that corresponding to a pres- 

 sure of from 210 Ibs. to 235 Ibs. per square inch. This, however, is 

 without taking account of the strain exerted longitudinally upon the, 

 shell of the boiler by the pressure on the ends, and it is upon the assump- 

 tion, which is hardly tenable, that the boiler heads, and especially the 

 Hues, are of the same strength as the cylindrical body or shell. With- 

 out the angle-iron strengthening recommended by Mr. Fairbairn, the 

 collapsing pressure of the flues of large boilers was found, in that gen- 

 tleman's experiments, to be sometimes as little as 87 Ibs. per square 

 inch. The strain resulting from the circumferential and longitudinal 

 components, in the outer shell, is one-eighth greater than that calcula- 

 ble for the circumference alone, so that, even if the heads and Hues 

 were stayed to the strength of the shell, this would correspond to a 

 pressure of but from 190 Ibs. to 210 Ibs., instead of 210 Ibs. to 235 

 Ib.s., as just supposed. But these estimates are for the strength of the 

 boiler when new. 



Corrosion the great Destroyer. In the experience of the officers 

 of the Manchester Boiler Association, with from 1,300 to 1,600 boil- 

 ers always under their care, one boiler out of every seven, and, in 

 some years, as in 18G2, nearly one of every four became defective by 

 corrosion alone, while of every eight boilers examined in the course of 

 a year seven are found to be defective in some respects. Thus, in 

 1802, with 1,370 boilers under inspection, 85 positively dangerous, 

 and^DS7 objectionable defects were discovered, 37 dangerous and 270 

 objectionable cases of corrosion alone having been reported. As a 

 boiler malady, corrosion corresponds in its comparative frequency and 

 fatality to the great destroyer of human life, consumption. It is the 

 one great disease. It is frequently internal, in consequence of tho 

 presence of acid in the water; but it is still oftener external, and it is 

 most insidious and certain wherever there is the least leakage of steam 

 into the brick-work setting. Condensed steam, or distilled water, is 

 an active solvent of iron, as well as of lead, and peaty water, which, so 

 far as inorganic matted- is concerned, is very pure, and distilled water 

 from surface condensers, and, indeed, any water that is quite soft is 

 known to eat rapidly into the substance of the boiler in which it is used. 

 A trickling, however slight, of condensed steam, down {lie outside of 

 a boiler, will infallibly cause corrosion, and to this was directly traced a 

 large number of the 47 boiler explosions which occurred in the United 

 Kingdom in 1803, 



Cast Iron not corroded. Corrosion is most rapid where the iron 

 is comparatively pure, as in the best Yorkshire and Staffordshire 

 plates. The presence, however, of a small proportion of carbon, as 

 in steel, or especially of silver and carbon, as in cast-iron, renders 



