Miscellaneous Papers. 387 



with the longest dimension in each case running across the grain 

 of the wood. In the one the medullary rays rao in direction of 

 the length of the piece and in the other at right angles to- it. To- 

 gether these were immersed in warm water, and when saturated 

 were taken out and measured. It was found that the total expan- 

 sion of the piece with the rays perpendicular to its length was fully 

 four times as great as in the one with the rays parallel. This test 

 being made with white oak this difference is probably abnormally 

 high. It is found that the ratio of difference prevailing in most of 

 our woods is about three to two. In this case the medullary ray 

 seems to be able to resist the natural forces resulting from the ex- 

 pansion or contraction. The question then arises, "Can the me- 

 dullary ray in a like degree resist such exte'rnal forces as may be 

 brought to bear upon the timber ?" 



In order to answer this question a number of tests were made 

 with various woods. The following are the principal tests made : 

 Longitudinal shear, deflection or cross- bending, resistance to cleav- 

 age and compression or crushing across the grain. 



In the longitudinal shear test one-half of the specimens were 

 prepared with mortises running through the stick parallel to the 

 pith rays, the other half with the mortises perpendicular to the 

 rays. The average result of all the pieces showed a small per cent, 

 of difference in favor of the ray. 



For the cross-bending test a piece of wood was sawed out so that 

 the rays would be exactly parallel to two faces and at right angles to 

 the other two. This stick was then sawed in two and each half tested 

 separately, the first with the ray parallel to the lines of force and 

 the second at right angles. The results of this test showed that 

 the piece with the rays presented edgewise to the load was able t(^ 

 support fifteen per cent, more for a given deflection, all other thing.s. 

 being equal, than the piece with the flat side of the ray to the load. 



In testing the resistance to cleavage, where the lines of force run 

 parallel to the ray the resistance was approximately 1500 pounds 

 to the square inch. With the opposite position the average was; 

 less than 500 pounds per square inch, and in each instance in the 

 latter tests the failure occurred in one of the rays, showing that to 

 at least one force the medullary ray yields before the fibro-vascular 

 tissue does. 



It was in the compression tests that the greatest difference was 

 manifest. For this test the blocks were cut to exactly the same 

 dimensions, the difference being that in one set the rays were verti- 

 cal and in the other set horizontal. They were taken one at a time 



