GENERAL METHODS 135 



provement on the simple cement filling, but its re- 

 sults were at best uncertain, and embarrassing 

 fractures continued to appear in fillings of any 

 considerable size. 



About this time (1907) the writer had to take 

 out a number of concrete fillings, some eight or 

 ten years old, which, though otherwise well done, 

 had been put in without previously treating in any 

 way the inner surface of the excavated cavity. 

 He observed, though hardly with surprise, that 

 while the concrete was perfectly hard and sound, 

 it was divided into sections, in every situation ex- 

 cept the very base of a grounded filling, by roughly 

 horizontal fractures. Many of these cracks were 

 so small that they could hardly be detected on 

 the surface of the concrete. These observations 

 made it quite apparent that it is idle to try 

 to increase the strength of a hollow trunk by filling 

 it with a rigid back-bone of concrete. And a little 

 consideration will make evident the soundness of 

 this conclusion. The secret of the great strength 

 of wood lies mainly in its elasticity, the ease with 

 which it can be bent and twisted. Concrete is 

 strong in quite a different way. It is almost per- 

 fectly rigid and inelastic. And it is futile to try 

 to reinforce an elastic substance with an inelastic 

 one. Suppose we have two beams, one of con- 

 crete and one of wood, bridging a certain space. 

 Suppose the concrete beam has a breaking point 



