In the preceding article, we introduced the theoretical framework of two-dimensional fluorescence lifetime correlation spectroscopy (2D FLCS). In this article, we report the experimental implementation of 2D FLCS. In this method, two-dimensional emission-delay correlation maps are constructed from the photon data obtained with the time-correlated single photon counting (TCSPC), and then they are converted to 2D lifetime correlation maps by the inverse Laplace transform. We develop a numerical method to realize reliable transformation, employing the maximum entropy method (MEM). We apply the developed actual 2D FLCS to two real systems, a dye mixture and a DNA hairpin. For the dye mixture, we show that 2D FLCS is experimentally feasible and that it can identify different species in an inhomogeneous sample without any prior knowledge. The application to the DNA hairpin demonstrates that 2D FLCS can disclose microsecond spontaneous dynamics of biological molecules in a visually comprehensible manner, through identifying species as unique lifetime distributions. A FRET pair is attached to the both ends of the DNA hairpin, and the different structures of the DNA hairpin are distinguished as different fluorescence lifetimes in 2D FLCS. By constructing the 2D correlation maps of the fluorescence lifetime of the FRET donor, the equilibrium dynamics between the open and the closed forms of the DNA hairpin is clearly observed as the appearance of the cross peaks between the corresponding fluorescence lifetimes. This equilibrium dynamics of the DNA hairpin is clearly separated from the acceptor-missing DNA that appears as an isolated diagonal peak in the 2D maps. The present study clearly shows that newly developed 2D FLCS can disclose spontaneous structural dynamics of biological molecules with microsecond time resolution.