An array of perfectly electric conducting (PEC) obstacles in a host dielectric medium modifies the constitutive parameters of the composite medium or metamaterial. Typically, for such a composite medium, the effective (homogenized) parameters are estimated using the Lorentz model. The Lorentz model indicates that inclusion of PEC obstacles in a host medium introduces both electric and magnetic polarizabilities that modify the permittivity and permeability tensors of the composite medium. On the other hand, the Floquet eigenmodal analysis shows that the PEC obstacles, while modifying the permittivity tensor, have no effect on the permeability tensor. Hence, as a corollary, Floquet model leads us to conclude that a negative permeability metamaterial using PEC obstacles is not possible to obtain. Interestingly, numerical results obtained from full wave analyses of metamaterial-slabs support the Floquet model. We will present the Lorentz and Floquet models and examine the source of inaccuracy in the Lorentz model. We will discuss the characteristic features of metamaterials made with non-resonant and resonant types of obstacles. Numerical results obtained from four independent models will be compared and discussed. In light of Maxwell’s equations, we will scrutinize the rationality of the claims of “negative-refraction” metamaterials. Co-sponsored by: IEEE AP-MTTS SBC IIT Kharagpur Speaker(s): Arun K. Bhattacharyya, Room: NKN Room, Bldg: E&ECE Department, Kharagpur, West Bengal, India, IIT Kharagpur, Kharagpur, West Bengal, India, 721302