The Strength of Materials as Applied to Engineering. 159 



tire strain in a direction perpendicular to their fibre ; I have 

 found that American white pine will not bear a strain greater 

 than five hundred pounds per square inch under these condi- 

 tions, and that even less than this will leave a permanent im- 

 pression ; no regard is however paid to this fact, and it is very 

 common to find premature rot produced in wall plates, bol- 

 sters and chords of bridges, by a partial crushing of the tim- 

 ber. The same is also true with regard to the bearing of 

 brace-blocks in bridges. 



A similar fact came to my notice. It was that of an ice 

 bouse. The ice, to the amount 200 tons, being packed on the 

 second story, the entire floor was constructed to bear its pro- 

 portional weight per square foot of the entire mass. The ice 

 was packed and supported safely until towards the close of the 

 first season, when the mass, by use and thawing, was reduced 

 to less than one-half, when suddenly the center portion of the 

 floor began to bend and finally to break, requiring a number 

 of props for safety. Before the packing of the second season, 

 additional joists were placed between the broken ones and a 

 straining beam under all, and suspended at intervals of nine 

 feet from trusses above ; still, at about the same time as in 

 the previous season, the centre interval broke. I was called 

 into consultation and examined the case carefully. I found 

 that, after admitting a loss of one-third the strength of the 

 timber from moisture, the original floor should have s£^fely 

 supported the entire weight desired, provided the same had 

 been uniformly distributed. This led me to re-examine the 

 matter. I found that the ice, which was a solid mass, had 

 melted beneath from the heat conducted by the zinc-covered 

 floor and assumed a convex form, so that the center joist, 

 which would break at sixteen tons, broke when overloaded, 

 and its companions in turn, until the spherical segment in 

 •contact rested upon a sufficient number of joist to support it. 

 This is not by any means a singular occurrence, but I doubt if 

 the same has been taken into consideration by builders. 



