350 Prof. Knoblauch on the Connexion between the 



group. In the second group this difference is less than in the 

 third ; and in the third, again, less than in the fourth (as with 

 with- and cross-grain pieces of poplar). When, therefore, the 

 fibres of all kinds of ivood are set in vibration, the pwity of reso- 

 nance is greater when such vibrations are transverse than when 

 they occur in other directioTis {as when the rods are cut across the 

 grain). But this superiority of resonance is not constant ; it de- 

 pends upon the nature of the wood. The difference in this respect 

 in the first group of woods is so small, that the resonance of two 

 with- and cross-grain pieces resembles that of two not very 

 dissimilar masses of stone when struck. In the last group the 

 difference is so great, that the tone of the with-grain piece when 

 struck has a metallic ring, while the dull sound of the cross- 

 grain piece reminds one of a piece of pasteboard when struck. 

 The division of the woods examined, derived from their thermo- 

 conductive power , is accordingly supported by their acoustic relations. 

 By supporting the two ends of the rods employed in the above 

 experiments and loading them equally in the middle, the degrees 

 of deflection which they undergo will give us an insight into 

 their structural relations ; for the greater their compactness, the 

 greater the resistance they will offer to bending ; and the less 

 compact they are, the more easily will they yield. The differ- 

 ence in vertical height of the middle points of the bent and 

 straight rods was taken as measure of deflection. A lever was 

 employed to determine this measure, the end of which passed 

 over an enlarged scale in order that the readings off might be 

 the more exact. The unit of this measure was a matter of in- 

 difference, inasmuch as in the comparison to be instituted, rela- 

 tions only had to be determined. Although, as was to be ex- 

 pected, in all cases the with-grain piece was much less flexible 

 than the corresponding cross-grain piece, yet an important dif- 

 ference was noticeable in the diftereut groups. This is best 

 seen by calculating the relation between the bending (measured 

 as above described) of the with-grain and that of the cross- 

 grain wood ; that is, the same weight being applied (say 100 

 grs.), by dividing the number given by the lever with the cross- 

 grain piece by that given with the with-grain piece. This rela- 

 tion (called " ratio of deflection " in the following Table) has, in 

 the first group, the mean value of 1 to 5 ; in the second, 1 to 8 ; 

 in the third, 1 to 9'5 ; in the fourth, 1 to 14. The division of 

 the groups is therefore also supported from this point of view*. 

 The difference in the structure in the different directions is least in 

 those woods which show the least difference with respect to direc- 



* The diversity of nature, even with one and the same kind of wood, of 

 course did not admit of the boundaries of the groups being drawn with 

 great exactness, or of the subdivision of the groups into secondary ones. 



