The formation and desorption processes of a self-assembled monolayer, deposited from solution, have been observed in situ, at the liquid/solid interface, using atomic force microscopy. The growth begins with the nucleation of submonolayer islands, that grow and eventually coalesce. The submonolayer island nucleation rate was found to depend on the concentration of the deposition solution. The power-law dependence of the submonolayer island density on the solution concentration was consistent with the interpretation that the smallest stable island consists of two molecules. By following the time-dependence of the island nucleation, and the growth kinetics of individual islands, we found that the formation process can be quantitatively explained by a kinetic theory of 2D cluster growth typically used to describe vapor phase molecular beam epitaxy. During the desorption process, holes in the monolayer are observed to nucleate, grow, and percolate across the sample, leaving isolated monolayer islands that gradually decrease in size. The relative rates of hole growth and hole nucleation suggest that removing a molecule from the monolayer/hole boundary is about 105 times more likely than removing a molecule from within a continuous region of monolayer. The coverage kinetics and nucleation and growth rates are consistent with a picture of monolayer desorption where molecules detach from the monolayer/hole boundaries, remaining adsorbed in the 'hole' regions from which they eventually desorb into the solvent / acase@tulane.edu

acase@tulane.edu

How do equity issuing firms and underwriters get together? We develop a theoretical model founded on the idea that issuers and underwriters associate by mutual choice. Underwriters look to the quality of the issuers who may wish to employ their services and issuers look to the abilities of the underwriters they consider employing. Our approach contrasts to the conventional view of one-sided choice established in the existing literature Our model suggests that a rematch will occur for subsequent offerings should changes in firm quality and/or underwriter reputation warrant it. Our empirical finding that firms experiencing a relative reduction in quality from IPO to SEO switch to lower reputation underwriters at the SEO stage provides especially convincing support for this notion We derive new implications about underwriter market share, showing that (a) high reputation underwriters benefit from a more stable revenue stream and (b) that while their market share increases when overall market activity is reduced, this would be accompanied by a relative decrease in the quality of the issues they underwrite In our model, pricing of underwriter services is left to bargaining between the parties after the match between issuers and underwriters is determined. While flat percentage IPO fees are a possible outcome with double-sided matching, high-reputation underwriters make higher dollar revenues by underwriting larger offerings. This is also true for subsequent offerings since high reputation underwriters associate with firms that undertake larger and more frequent security offerings subsequent to their IPO We also find that underwriter reputation is negatively related to the probability to withdraw an offering during the registration period and that on average in adverse market conditions underwriter reputation has a higher negative impact on the probability to withdraw than it has in more favorable market conditions By examining gross spreads over time we find a significant clustering of percentage gross spreads on round numbers. The evidence suggests that while fixed costs of issuing equity have not changed over time, variable costs have increased. At the same time, percentage spreads have become less sensitive to the relative size and the risk characteristics of the issue / acase@tulane.edu

acase@tulane.edu

This dissertation report's calculations of ionization-excitation cross sections of helium and, for the first time, of the negative hydrogen ion, differential in the energy transfer and total, by Compton scattering. They are compared to the Compton cross section by a free electron. The electronic initial state is the ground state, used with two limits, uncorrelated and fully correlated. The final state is selected as uncorrelated. We find the H-- cross section, when the bound electron remains in the ground state, to be 20% smaller than He. This was not expected since H -- geometrical cross section is about 16 times larger than He. However when the sum over the bound states of the second electron is taken into account both cross sections are comparable with two times the Compton cross section by a free electron. A general analysis based on the use of sum rules and the quasi-elastic scattering is applied to both shake and correlated calculations of the cross sections / acase@tulane.edu

acase@tulane.edu

acase@tulane.edu

acase@tulane.edu

acase@tulane.edu

Deterministic automata are induced from sample strings using a population based algorithm. Deterministic finite state automata are induced from regular languages, deterministic pushdown automata are induced from context free languages, and deterministic Turing machines are induced from regular languages, context free languages, and context sensitive languages. The algorithm is a variation of genetic algorithms using tree shaped chromosomes to represent the transition tables of the various automata. When Turing machines are induced, they are encapsulated and are available to subsequent inductions as primitive operations. Turing machine schemas produced this way may be nested arbitrarily deep / acase@tulane.edu