Presumably those interested in methods of testing building constructions need not normally be concerned with the strength of the individual pieces that make up a building unit. Such disinterest is understandable when the strength of the individual pieces has been previously established and standardized. However, if strength properties have not been established, as in the case of nonstress graded lumber, the results of the tests of a building construction might well be questioned. For example: Light trussed rafters, spaced at close intervals such as 16 or 24 in. on centers in residential type buildings, are rapidly increasing in use. They are manufactured in factories or built at the job site. Many producers have their own truss designs, together with a proprietary system of joint connections. Proving adequacy of these trusses by testing seems to be the usual method of gaining building official and Federal Housing Administration acceptance of the design. Usually the structural members comprising these trusses are 2 by 4 in. pieces, and most of the 2 by 4 in. lumber that is readily available is yard lumber that is not stress graded. This often creates a problem in obtaining the desired approval, even though the tests of these truss assemblies show unusually good strength characteristics. Such a difficulty could be expected, for it can be argued that the excellent test results were obtained by the coincidence of better than average lumber quality in the frame. It is only one example of how the unknown strength quality affects the usefulness of this class of lumber.