In part 1 of this series, we proposed a new method to determine the correct equilibrium melting temperature (T-m(0)). Effects of the "melting kinetics" and lamellar thickening were omitted from T-m. The correct T-m(0) of isotactic polypropylene (iPP) ([mmmm] = 99.6%, M-n = 64 x 10(3), and M-w/M-n = 2.4) was observed to be 186.1 degreesC. In this paper, the rigorous Gibbs-Thomson plot was obtained by using the direct correspondence between maximum melting temperature (T-m,T-max) and maximum lamellar thickness (l(max))- T-m,T-max and l(max) were observed by means of optical microscope and transmission electron microscope (TEM), respectively. The validity of the Gibbs-Thomson plot obtained by means of a differential scanning calorimeter (DSC) (part 1 of this series) was confirmed by comparing it with the rigorous Gibbs-Thomson plot in this paper. The Hoffman-Weeks plot is widely used as one of the methods to obtain T-m(0). It was shown that the Hoffman-Weeks plot is correct only when l proportional to 1/DeltaT where DeltaT is the degree of supercooling, is satisfied. However, in the case of iPP, the condition is not satisfied, and so the result obtained by the Gibbs-Thomson plot is not equivalent to that obtained by the Hoffman-Weeks plot. The existence of alpha2' phase was confirmed again by breakings in slopes of l and T-m against T-c at 159 degreesC. Furthermore, the broad bimodal distribution of l was caused by the difference between the lamellar thickening growth rate of isolated mother lamellae and the lamellae thickening rate of stacked daughter lamellae.