X-ray bursts are some of the most energetic events known, releasing roughly 1039 ergs in 10 to 100 seconds. The energy for these events is thought to be provided by the explosive thermonuclear runaway occurring near the surface of accreting neutron stars. During the explosive event, the γp-process, which is characterized by a series of (γ,p) and (p,ã) reactions, may play an important role in determining the time-dependent release of energy. Models of X-ray bursts are used to test predictions of the physical characteristics of the neutron star surface, such as the temperature and density, as well as to predict the distribution of nuclei created during the burst. To determine the time dependence of the energy release for the models, nuclear reaction network calculations are performed. These calculations are based on the thermonuclear reaction rates of the reactions taking place during the X-ray burst, therefore the reliability of the burst models is dependent on the accuracy of the reaction rates.   To investigate the importance of the γp-process, the 32S(p,t)30S, 36Ar(p,t)34Ar, and 40Ca(p,t)38Ca reactions were studied to determine the excitation energies of the excited states in 30S, 34Ar and 38Ca above the α-thresholds with uncertainties below 10 keV. The properties of these excited states determine the (γ,p) reaction rates. The experiments were performed at RCNP in May 2008 using 100 MeV, dispersion matched protons and the Grand Raiden Spectrometer. Thin natural Sulfur and Calcium foils, each about 3.5 mg/cm2 thick, were used as targets. The Sulfur target was enclosed in thin Gold layers to keep the target from sublimating while being bombarded with beam. The 36Ar target utilized the gas cell designed at RCNP to be used with dispersion matched beam and highly enriched 36Ar gas. The 1 mg/cm2 thick gas target was contained in the cell by windows made of 6 ÌÂm Aramid foil. This experiment represents the first time a dispersion matched (p,t) experiment was performed on an extended target.   The high precision (p,t) data was collected at the focal plane using the two Multi-Wire Drift Chambers, capable of determining horizontal and vertical positions and angles of each ion detected. Additionally, two plastic scintillation detectors provided timing and energy loss information, which was used for particle identification. 24Mg(p,t)22Mg data was used to calibrate the focal plane and an asymmetric slit was used to perform the angular calibration. The high resolution triton data allowed for the full reconstruction of the reaction kinematics at the target and the energy of the states in the residual nucleus could be determined with uncertainties below 10 keV.   Of the 122 states observed in these experiments, 59 were seen for the first time. For the case of 30S, these mark the first states observed above the γ-threshold. The excitation energies determined here have been used to estimate the astrophysical reaction rates for the 26Si(α,p), 30S(α,p) and 34Ar(α,p) reactions. The present results indicate a smaller rate compared to the statistically based Hauser Feshbach rates currently being used in X-ray burst models. The lower rates imply it is possible the nuclear material proceeds through the mass 26 to 38 range more slowly, thus affecting the time distribution of the energy generated during an X-ray burst. With more accurate reaction rates, X-ray burst models are able to more reliably determine the characteristics of the physical environment near the surface of neutron stars.