174 ANNUAL OF SCIENTIFIC DISCOVERY. 



integrity of the iron work used in the construction. The method 

 to which we refer was discovered by Mr. S. M.'Saxby, R.N., and 

 has been tested upon a great variety of forgings at the royal dock- 

 yards at Sheerness and Chatham. The process depends upon the 

 very simple principle that when a bar or mass of soft homoge- 

 neous iron, of the best quality, and free from any flaws or defects 

 causing a separation of the particles of the iron, is placed in the 

 position of the dipping-needle, it is at once sensibly magnetic, 

 the lower end being a north pole in northern latitudes, and the 

 upper end a south pole. The same action takes place in a bar 

 hanging vertically, or in any other position, but to a less extent. 

 With internal flaws, however, the bar is no longer one regular 

 magnet, but several different magnets, with the different magnet- 

 isms separated from each other. When a delicately balanced 

 magnetic needle is passed over the bar to be examined, the bar 

 beinff placed east and west in the equatorial magnetic plane, if 

 the bar is entirely sound the needle remains at right angles with 

 the bar, that is, N. and S. ; but the moment a flaw, or a separa- 

 tion between the particles of the iron, exists, the needle departs 

 from its normal position and assumes a new direction, thus show- 

 ing the place of the fault. A large number of trials at the shops 

 of the royal dockyards were made in the presence of many engi- 

 neers and iron-workers, and the place of the fault indicated by a 

 chalk-mark ; afterwards the bars were broken in a testing ma- 

 chine, and the result in every case showed the decision of the 

 magnet to be entirely correct. In one case a round bar 14 inches 

 long had a hole drilled into it, and a bolt of unmagnetized steel 

 was inserted and welded up with the end of the iron bar. The 

 needle detected the fact that a fault existed at the place where the 

 bolt was put in ; and upon cutting up the bar it was found that 

 one end of the bolt was not welded to the iron. 



A large paddle crank-shaft was decided by the magnet to be de- 

 fective near one neck ; and upon turning the metal down the defect 

 was found. A bar was decided by*the test to be bad, and it was 

 afterwards found to have been made partly of good and partly of 

 bad iron. Another bar was declared faulty, and was found to be 

 solid, but " upset" in the middle of its length, and then hammered 

 down to its original diameter at a temperature below welding 

 heat. 



The experiments upon rolled plates, upon steel, and cast iron, 

 as far as made, are satisfactory, though farther examination is 

 needed to make the test practically available in this respect. Per- 

 haps the most valuable result of this mode of testing iron is the 

 detecting of that change called crystallization, which is so often 

 seen in iron, and is so important a matter to be regarded in the 

 use of railway axles. 



SOUK; of the experiments seem to show that the ultimate tensile 

 strength of a bar is greatest when the anvil upon which it is 

 forged stands in an east and west position, but that the highest 

 elastic limit is attained when the anvil stands in a north and south 

 position. American Hallway Times. 



