In this article, some recent work is surveyed on mechanical properties of both metals and semiconducting silicon. In quantum chemical language, such properties as (a) cleavage forces and (b) tribological properties involve bond stretching, and here it is argued that electron correlation must play an important role. In the context of long-range dispersion forces, this is already evident in first-principles work of Lifshitz, which, particularly in simple metals, leads to an asymptotic form of cleavage force F(z), with z the interplanar separation beyond equilibrium spacing on cleavage, falling off as z -3 , the magnitude of this term being expressible in terms of the electron plasma frequency. Here, it is argued, from quantum chemistry, that in covalently binded systems, such as semiconducting Si, the elasticity of a covalent bond in the crystal is quite comparable with that of the H 2 molecule bonding in free space. The transition from delocalized molecular-orbital theory to Heitler-London or valence bond regimes comes about as the bonds are stretched to breaking point. This situation with covalent bonds is compared and contrasted with that of a largely nondirectionally bonded metal such as Cu, in which there is a closed d shell, and a more complicated body-centered-cubic Fe, in which the charge density can be demonstrated experimentally, through "overlapping" reflections in X-ray diffraction, to have marked directionality. Even for simple sp metals, it is shown that some of the main features of the heavier alkalis, and in particular K, Rb, and Cs, can be explained by viewing the dimer, say K 2 , as the basic building block in representing the energy of the crystal of K as a function of lattice spacing and local coordination number. Finally, a brief discussion of liquid alkalis is given, mechanical properties now embracing viscosities.