Since Lick indices were introduced in 1994, they have been used as a source of observational data against which computer models of galaxy evolution have been compared. However, as this thesis demonstrates, observed Lick indices lead to mathematical ill-conditioning: small variations in observations can lead to very large differences in population synthesis models attempting to recreate the observed values. As such, limited reliance should be placed on any results currently or historically in the literature purporting to give the star formation history of a galaxy, or group of galaxies, where this is deduced from Lick observations taken from a single instrument, without separate verification from at least one other source. Within these limitations, this thesis also constrains the star formation histories of 21 nearby elliptical galaxies, finding that they formed 13.26 +0.09 -0.06 Gyrs ago, that all mergers are dry, and that galactic winds are formed from AGN activity (rather than being supernovae-driven). This thesis also finds evidence to support the established galaxy-formation theory of "downsizing". An existing galactic model from the literature is examined and evaluated, and the reasons for it being unable to establish star formation histories of individual galaxies are ascertained. A brand-new model is designed, developed, tested and used with two separate data sets, corroborated for 10 galaxies by data from a third source, and compared to results from a Single Stellar Population model from the literature, to model the star formation histories of nearby elliptical galaxies.