446 THE STRENGTH 



It is evident that every cross section of this 

 model had the same quantity of matter in it, 

 and might represent that of a beam with any 

 proportion of flanges. From this model I got 

 castings made, and fixing them firmly at one 

 end in a horizontal position, broke them by 

 w^eights hung at a certain distance from that 

 end, the narrow^er flange being always at the 

 bottom or the compressed one. 



The following five experiments were made on 

 as good a casting as could be procured ; it was 

 not quite uniform, but balanced when laid 

 across a thin surface, at ~ of an inch from the 

 middle. Its whole length was 7 feet 8 inches. 



Experiment First — Arm 4 feet. 

 Breaking weight, including the pres- j 



sure from weight of casting at J = G65 lbs. 

 place where weights were hung, j 

 Ratio of flanges, at place of fracture, 3 to 13, 

 It broke one inch from the shoulder. 



Experiment Second — Arm 4 feet. 

 In this the fixed end of the casting was changed, 



but otherwise the experiment was as before. 

 Breaking weight, including weight ■^_^. 



^ 1 I — O I *7 IDS. 



from arm as above, > 



Ratio of flanges, 7 to 23. 

 It broke 3| inches from the shoulder. 



