Stellar abundances are essential parameters for many branches of astronomy. They are also particularly important as tests for theories of nucleosynthesis. Despite the encouraging results from band photometry, detailed spectroscopic analyses are still important in furthering knowledge of physical processes occurring in a variety of stellar objects. This thesis presents the first of a series of absolute abundances for the late-type giant Arcturus. The broader aim of the investigation is to make Arcturus a reliable standard star, and also it is hoped that equilibrium assumptions in its atmosphere can be scrutinised. An essential prerequisite to a model atmosphere is the continuum flux curve. Certain discrepancies near the flux maximum prompted the construction of a spectrum scanner which was used to obtain fluxes of Arcturus at the Wise Observatory, Israel. All known flux results have been collated for Arcturus, and the continuum flux curve is presented for the region 3800 andAring; to 13 microns. Ways of measuring the effective temperature are discussed. The only adequate method for Arcturus and other cool stars involves comparing the relative energy distribution with model predictions. As all present models fail to account for the flux below 5000 andAring; this method is not as accurate as it could be. A value of 4450 ± 50°K is adopted for Arcturus; most of this error is due to the unrealistic nature of the models. Using this temperature a grid of line-blanketed models with different gravities has been constructed. The suggested gravity-indicator requires a precise value for the microturbulence, and for this preliminary analysis a model with log g = 1.7 was selected. This value, reported by several investigators, is later shown to be fairly realistic. Uncertain oscillator strengths continue to plague abundance work in astronomy, though the Oxford furnace is now producing precision values for several elements. Using this furnace a series of neutral titanium lines of direct astrophysical interest have been measured, and they are accurate to at least ±0.03 dex(7%) on a relative scale. Considerable effort is still needed in establishing accurate absolute scales however; these results have been made absolute by referring to theoretical and observational lifetimes. The internal consistency of the Ti I values has been verified by studying the solar abundance of titanium. Rf restricting the analysis to good quality weak lines, uncertainties in damping and microturbulence have been avoided. A log abundance of 4.81 ± 0.08 is reported (on the log H = 12.00 scale). For the Arcturus analysis, a few iron lines have also been included. In this star the abundances are very sensitive to the value assumed for the microturbulence. The empirical nature of this parameter is emphasised. A value of 2.1 ± 0.2 km sec^-1 was found by demanding an unique abundance for all lines. Theories accounting for the origin of iron-peak elements are discussed. Explosive Silicon burning cannot account for all the observed isotopes, and the equilibrium process has been re-introduced to explain some anomalies. These theories are examined in relation to supernova models and Galactic evolution, and also with regard to the broader purpose and prospects for measurements of stellar abundances.