108 M. Savarfs Researches on the structure of Metals^ 



the other of two branches of a hyperbola ; but with this pe- 

 cuUarity, that one of the nodal lines of the rectangular system 

 is always placed on the line which bisects the angle which the 

 fibres of wood form with each other, and that one of the 

 asymptotes of the hyperbolic curve appears to be sensibly 

 parallel to the direction of the fibres of one of the plates, while 

 the second is parallel to the fibres of the other plate. We 

 shall obtain perfectly analogous results, by crossing any two 

 plates which contain at least one of the axes of elasticity, that 

 is, in which one of the nodal systems are formed by two lines 

 at right angles to each other. If one of the two plates does 

 contain any of the axes in its plane, then the nodal systems are 

 composed only of branches of a hyperbola, and the position 

 which they take is intermediate to that which they affect in 

 each of the plates considered separately. We may therefore 

 conclude, that, in whatever manner bodies which possess these 

 rectangular and unequal axes of elasticity are united together, 

 their effect when combined, is to exhibit all their axes of elas- 

 ticity. 



In general there does not appear to be a great difference 

 between the structure of plates of metal which have been cut 

 out of great masses, and that of plates of the same substance 

 which have been melted in moulds to give them the circular 

 form. Among both these are found some in which the inter- 

 val between the two sounds is only very small, while in others 

 it embraces several tones. In the plates formed in moulds it 

 is remarkable, that the substance of the mould, the position 

 of the jet at the circumference or at the centre, the direction 

 of the mould, whether vertical, horizontal, or inclined, do not 

 appear to have any influence over their state of elasticity ; 

 that is, we find always a direction of the greatest resistance to 

 flexion, as well as the two modes of division affecting determi- 

 nate positions and accompanied by different sounds. 



It does not appear that a sudden cooling, nor an electric cur- 

 rent which traverses the plate in one of its diametral lines 

 while the metal is in fusion, exercise any appreciable influence 

 over the general character of the phenomenon ; but it is other- 

 wise with a series of small blows given to the mould whilst the 

 metal is becoming solid, as this last action almost never fails 



