27 



for any one tree, given in the last column of Table HI, represents an average percentage of all 

 finalities, or it may be considered an indication of the relative average quality. These averages 

 have been used in iilotting the acconipanyiug diagram, in the order of their magnitude. No 

 attempt is made here to explain the apparent discrepancy of 20 per cent between the lowest and 

 highest average quality, any further than to call attention to the fact that in a general way this 

 variation in strength corresponds fairly well with the variation in specilic gravity. 



COMPAKISON OF SINGLE QUALITIES WITH THE AVERAGE QUALITY. 



While the relations which exist between the relative average quality of a tree and its relative 

 standing in specific gravity, compressive and cross-breaking strength (exhibited in Fig. 14), all 

 agree, in a general way,with those trees which give less values in one of these directions having less 



120 % 



Fig. 14. Relative average qu.ality of different trees aa compared with strength in cros8-hreaking and crushing 



endwise. Long-h^afpine {Pinna ^'o^i'^f'''")- 



values in all the others here named. In some other matters the agreement is not so close, as may be 

 found by plotting the corresponding percentages given in Table III upon this figure. It has been 

 well established that strength is no function of the width of the annual rings. It is, however, a 

 function of the jn-oportion of summer wood to spring wood in each annual ring, or, since the sum- 

 mer wood is always more dense, this is tlie same as saying that the strength is a function of the 

 density or of the specific gravity, which relation is clearly shown in Fig. 14, as well as in Plate xi. 



RELATIVE STRENGTH OF LARGE AND SMALL ISEAMS. 



In Table IV the mean results of all tests on large and small beams are tabulated separately. 

 By comparing these mean values, as given at the bottom of Table IV, we are forced to conclude — 



(1) That large beams are from 10 to 20 per cent weaker in ultimate strength than the corre- 

 si)onding 4x4 inch beams from the same logs. 



(2) That large beams are stronger at their elastic limits than the small beams from the same 

 logs. 



