[ 1* ] 



IV. Experiments on the Cohesion of Cast-Iron. By B. Bevan, 

 Esq. 



To Richard Taylor; Esq. 



Sir, 



A N erratum, of a single figure in the third formula in the 



-^ last Number of the Philosophical Magazine, page 343, 



may have a tendency to mislead some of your readers. It will 



be noticed that the third as printed, is the same as the second, 



but should have been , , = c. 



a- 



Experiments on the cohesion oi cast-iron lead to very irre- 

 gular results, depending in some degree upon the quality of 

 the metal ; but in a greater degree upon the improper mode 

 of applying the force. During the present year I had some 

 cylmders and prisms cast of gray soft metal ; I then reduced 

 the middle part of one of the cylinders to '425 inches diameter 

 and fixed it in my experimental press, with the resultant of 

 the straining force, nearly in the axis of the cylinder, the 

 weight required to break it was 2550 pounds; at the place of 

 fracture was a visible Jault in the casting, so that the proper 

 strength must be more than 17,900 pounds to the square 

 inch. 



I then took another cylinder and reduced it, by turning, to 

 exactly '5 or ^ an inch in diameter, and submitted it to the 

 like process. The breaking weight was 6430 pounds, showing 

 a strength of cohesion of 32,700 pounds per square inch. 



This last specimen sustained a load of 5988 pounds for 

 five minutes, before the last motion or addition of the weight : 

 so that we cannot be wrong in estimating iron of this quality 

 to have a cohesion of more than 30,000 pounds to the square 

 inch. The specific gravity of the iron was 7"7 16. I also tried 

 the transverse strc7igth of some of the same iron cast into 

 prismatic bars : these bars were placed horizontally upon two 

 supports, twelve inches asunder, and the weight applied on 

 the middle of the bar ; the depth being '65, and breadth '49 

 inches respectively. The gradual application of the load oc- 

 cupied three hours ; and this bar sustained a load of 700 

 pounds for ten minutes without signs of fracture : an addition 

 of 10 pounds broke it. If we therefore take the formula 



= c, tlie cohesion appears equal to 61,000 pounds to 



6d2 



the square inch. ? 



Another bar of the same metal, and of nearly the same di- 

 mensions, was tried with the least thickness vertical, i. e. in 

 the manner of using it, the de})th and breadth being '487 and 

 '64, having had a part of the hardened surface removed by a 

 fine file. Being 



