394 CHAIN BRIDGE, 



ADDITIONS. 



In the preceding paper, we have supposed 

 the iron to be of the mean strength of sound 

 metal ; but if it should have any unseen cracks 

 or bad weldings in it, there is no saying 

 what would be an adequate compensation for 

 them, for its strength might be reduced in any 

 degree. In the specimens I have seen in broken 

 chains, some of which have been of great 

 dimensions, most of the failures have been 

 through defective welding; hence the necessity 

 of testing all chains, the consequences of whose 

 failure would be serious: they ought I think to 

 be tried with a high test, for if a chain were tried 

 with 9 tons per square inch, there is no proof, 

 further than the probabihty the defect would 

 have shown itself, that it would bear 12 tons, 

 much less 27. 



There may, however, be urged two objections 

 against high tests : 1st. The difficulty of applying 

 such immense strains, and which it is probable 

 leaves some bridges without their chains being 

 tested at all: 2nd. That high tests injure the 

 elasticity of the metal, and consequently impair 

 its strength. 



