6o Mayow 



be drawn downwards towards the rod. But indeed 

 since the rigid body is as solid and compact in respect 

 to its external parts as the more slender rigid body or 

 even more so (for the exterior parts of a thick rigid 

 body, be it glass or iron, are more quickly cooled than 

 the interior, so that it is extremely solid as respects 

 its external surfaces — in so far as the nitro-aerial 

 particles in endeavouring to escape are detained in 

 the exterior parts which have now cooled down, and 

 are fixed there as we have already shown — but this 

 does not take place in a more slender rigid body, since 

 all its parts are cooled at nearly the same moment), 

 hence I say it results that the convex surface of a 

 thicker rigid body cannot bear to be drawn out as far 

 as is necessary for bending it without compression of 

 its parts, so that it is now necessary that the convex 

 surface of the shortened rigid body should also go 

 inwards and make, say the Hne ^, m^ h, in the figure ; 

 and that cannot be done without notable compression. 

 And the further the two surfaces are from one another, 

 so much the more must the convex surface pass 

 inward during the bending, and consequently the 

 matter of the rigid body will suffer the greater com- 

 pression ; so that very thick rigid bodies cannot be 

 bent. And thus it is that a broad and thin plate is 

 easily bent so far as regards the broad surfaces which 

 are near each other, while as regards the lateral sur- 

 faces which are much farther apart it cannot be bent. 

 While we thus maintain that the power of recoil in 

 rigid bodies should be ascribed to the compression of 

 their matter, I would not be understood as thinking 

 that matter thus compressed endeavoured to extend 

 itself, for that would be to assume elasticity but not to 

 explain it, and any one would be ready to ask : 

 whence arises the power of recoil in the compressed 



