Comment on ''Quantummoleculardynamics simulations of liquid metals and highly degenerate plasmas'' [Phys.Molecular dynamics simulation of water beween two ideal classical metal walls Many alkali-post-transition group IV alloy systems exhibit clearly defined equiatomic compounds together with a pronounced intermediate range ordering, indicated by a first sharp diffraction peak at Ϸ0.9 Å Ϫ1 . These phenomena have been explained assuming that tetrahedral group IV anions, ''Zintl'' ions, survive in the liquid state. As a prototype system we considered liquid CsPb, for which several experimental results are available, and studied it by means of first-principles molecular-dynamics. Agreement with experiment is satisfactory, provided the 5s and 5 p electrons of cesium are explicitly taken into account in the computation of the electronic valence charge density. In particular, our calculations reproduce the structure factor prepeak reasonably well. The local liquid structure however is quite complex. This can be described as a disordered network, which still has many features in common with the ''Zintl'' ion model. For instance, the average Pb-Pb coordination is close to 3, the value for perfect tetrahedra, but the coordination distribution of Pb around Pb shows a broad range of values. The calculated electronic density of states shows a minimum at the Fermi level indicating compound formation.