In a recent paper, we represented the partial polymer polarizability tensor equation in terms of partial monomer polarizability tensors and intermonomer unit dipole-dipole interaction tensors. Here, we represent the same equation in terms of a partial polymer Green's function matrix equation being constructed from partial monomer Green's matrix elements and intermonomer interaction energies. We also show that the circular dichroism (CD) and ultraviolet (UV) absorption band shape functions bear a strong resemblance to DeVoe's scalar representation of the band shape functions, which we have rewritten, using our terminology, in terms of the three terms describing the dipole-dipole coupling, the electromagnetic coupling, and the intrinsic optical activity. The latter is capable of incorporating the one-electron mechanism. The present tensor theory offers five advantages. First, for describing physical phenomena, the tensor form may be more convenient than the scalar forms. Second, in both the partial polarizability tensor theory and DeVoe's partial scalar polarizability theory, it is possible to place different transition dipoles within each monomer in different positions, whereas this is not possible in the usual Fano-DeVoe types of polarizability theories. Third, the intermonomer interaction energy can be represented readily either by the dipole approximation or by the monopole approximation. Fourth, various types of intermonomer interactions can be taken into account flexibly by making use of the Green's matrix equation. Fifth, an efficient algorithm can be derived by reducing the matrix size by the factor 3 x 3 as compared with the partial polymer polarizability matrix tensor equation. By employing the dipole positions already known or discussed by other authors for the two n pi* and pi pi* transitions within each amide monomer, the theory has been applied to the semiempirical calculations of the CD and UV absorption band shapes of a cyclic pentapeptide, cycle (Gly-Pro-Gly-D-Ala-Pro), with comments on the adaptability of simplified parameterization.