58 Mayow 



trude it towards the sides. Whence it results that 

 as rigid bodies on bending are attenuated as to their 

 thickness, so on the other hand they are somewhat in- 

 creased in breadth. 



Since in bending rigid bodies the convex side comes 

 thus to be lengthened and the concave to be shortened, 

 the result is that the thinner rigid bodies are, the more 

 and the more easily can they be bent, for although glass 

 is very fragile and can scarcely be bent, yet fine threads 

 of it can be wound round a bobbin and tied in a knot. 

 But that the reason of this difference maybe understood, 

 let ^, c, 3, d^ in Plate I., Fig. 9, be a very slender rigid 

 body whose convex and concave surfaces were equal 

 before inflexion — but now that it is bent, the con- 

 vex surface 3, d^ is a little elongated. Let us suppose 

 that two points e^ e^ are so placed in the convex surface 

 that the line between the limits ^, e^ is equal to the con- 

 cave surface ^, c^ which we suppose to be equal to the 

 length of the rigid body before inflexion. But now if 

 the convex surface of the bent rigid body be lengthened 

 out at both ends, at one end from ^ to 5 at the other 

 from eto ddiS is done in the figure, then there is no 

 need for this surface being drawn inwardly when the 

 rigid body is bent — nor consequently that its matter 

 should be compressed. And yet this is inevitable in 

 rigid bodies whose surfaces cannot be elongated, as we 

 have already shown. 



Further, if we suppose also that the concave surface 

 of the rigid body ^, ^, is shortened as much proportion- 

 ally as the convex surface is lengthened (for it should 

 be observed that the force by which a rigid body is 

 bent tends as has been shown above not only to 

 draw out the convex surface but also to contract 

 the concave) — say that the concave surface at each 

 end is contracted to «, and the convex lengthened 



