152 W. H. WARREN. 



deformations before fracture. Moreover, cracks appear 

 on the outer face long before actual fracture, thus giving 

 warning of approaching danger. 



The experiments also suggest that in order to obtain the 

 maximum strength for the minimum cost in reinforced 

 concrete columns, such as would be most suitable for fire 

 proof buildings, the reinforcing should consist of soft iron 

 or steel spirals having a longitudinal pitcJi of about -§- of 

 the diameter, with longitudinal steel rods arranged on the 

 inner side of the spirals, around the circumference of, say 

 Bessemer or other steel having a high elastic limit. It has 

 been shown by Oonsidere that such a combination is 

 approximately equal in resistance to riveted steel of the 

 same weight. It lias been suggested to submit such 

 columns to a test load sufficient to crush the outer shell 

 of concrete, and after the test to put on a concrete coat- 

 ing in which asbestos is substituted for the sand. 



Coefficient of elasticity. — The curves Fig. IV. m. and 

 Fig. IV. k. show that the coefficient decreases rapidly 

 with the increase in the stress, when the loads are gradu- 

 ally increased to tbe breaking point, but if the loads are 

 removed and reapplied the effect of the load first applied is 

 to increase the coefficient for the second load. 4 Fig. IV. m., 

 5 Fig. IV. m., and 5 Fig. IV. k. illustrate this point. The 

 application of the first pressure on a hooped concrete prism 

 raises the elastic limit. The coefficient of elasticity 

 obtained by repeated loading and unloading does not 

 diminish to the same extent as when the loads are gradually 

 increased to the breaking point. 



Transverse strength. — The experiments on the trans- 

 verse strength were made on mortar beams consisting of 

 1 of cement to 3 of river sand. These beams were 4 by 4 

 inches in cross section, tested on a span of 40 inches ; two 

 beams were reinforced with three rods \ inch in diameter 



