The systematics of the eastern North American isopods of the genus Asellus are revised, based on all available material (8918 specimens in 998 collections) and literature relating to eastern North American forms. Sixty-eight species (both epigean and hypogean) are recognized and discussed, five of which are newly described. Evidence is presented for synonymizing five nominal species of Asellus. A discussion is presented as to the generic status of the eastern North American isopods placed in the genus Asellus with the conclusion being reached that the proposed fragmentation of the genus is invalid (on the basis of the characters utilized). The existing species groups of troglobitic asellids have been modified and new species groups proposed for the epigean forms. The distributions of the various species groups and their individual species are given. The eastern North American isopods of the genus Asellus originated through multiple invasions of fresh water habitats primarily from early to middle Cenozoic during inundation of the coastal plains by a shallow marginal sea. Inland dispersal of these epigean ancestral stocks occurred by migration through then existing streams and rivers assisted most importantly, by the process of stream captures. Evolution of hypogean asellid stocks from epigean ancestral lineages is hypothesized. Subsequent patterns of speciation of troglobitic forms is discussed with consideration given to hypothetical barriers to dispersal. / Ph. D.

When beryllium is thermally diffused into silicon, it gives rise to acceptor levels 191 meV and 145 meV above the valence band. Quenching and annealing studies indicate that the 145 meV level is due to a more complex beryllium configuration than the 191 meV level. When lithium is thermally diffused into a beryllium doped silicon sample it produces two new acceptor levels at 106 meV and 81 meV. Quenching and annealing studies indicate that these new levels are due to lithium forming a complex with the defects responsible for the 191 meV and 145 meV beryllium levels, respectively. Electrical measurements imply that the lithium impurity ions are physically close to the beryllium impurity atoms. The ground state of the 106 meV beryllium-lithium level is split into two levels, presumably by internal strains. Tentative models are proposed to explain these results. / Ph. D.

Forty-four strains (including 20 type strains) of spirilla, have been subjected to a comprehensive nutritional characterization. A total of 67 different organic compounds were tested for their ability to serve as sources of carbon and energy. A total of 30 different compounds were tested for their ability to serve as sole nitrogen sources. Other characters that were studied included: urease production; NaCl tolerance of fresh-water Spirillum strains; effects of calcium ions on growth of Spirillum strains; production of fluorescent pigments; enzyme assays of key enzymes of carbohydrate catabolism in P. lunatum, A. itersonii, and A. peregrinum; and the effects of various inhibitors on glucose-6-phosphate dehydrogenase activity from P. lunatum. The resultant data, plus that of Wells (1970) and McElroy (1970), have been used to divide genus Spirillum into four genera (Aquaspirillum, Spirillum, Oceanospirillum, and Pseudospirillum), each possessing its own unique characteristics. Diagnostic keys are presented for the four genera and also the species of each genus. Each of the four genera were compared with other related genera of bacteria. The species O. minutulum (formally Spirillum minutulum) and O. halophilum (formally Spirillum halophilum) were combined into a single species (O. minutulum) on the basis of their morphological, physiological, nutritional, serological, and DNA base composition similarities. The two species O. linum (formally Spirillum linum) and O. atlanticum (formally Spirillum atlanticum), were also combined into a single species (O. atlanticum) because of their similarities. Previously unnamed strains 101, 102, and 206 were proposed as a new species, O. maris. With regard to the genus Aquaspirillum, one new species was proposed, A. mobile (formally unnamed strains 103 and 104). Characteristic enzymes of the Entner-Doudoroff Pathway were found in P. lunatum and A. itersonii. Key enzymes of the EmbdenMeyerhof-Parnas Pathway were detected in A. itersonii and A. peregrinum. The characteristic enzymes of the hexose monophosphate pentose pathway were not detected in any of the above species. Glucose-6- phosphate dehydrogenase from P. lunatum was found to be inhibited by ATP and also to be non-specific for NAD⁺ or NADP⁺. / Ph. D.

Two glutamate dehydrogenase isozymes were shown to exist in Chlorella pyrenoidosa (strain 7-11-05) and the cellular level of these isozymes was regulated by the nitrogen source upon which the algae were cultured. One isozyme was induced by ammonium and was specific for the coenzyme, triphosphopyridine nucleotide (TPNH). The isozyme present in cells cultured in nitrate-containing medium has a much higher specificity for the coenzyme, diphosphopyridine nucleotide (DPNH). The TPNH:DPNH activity ratios for the respective isozymes were 33:1 and 1:5 and their respective molecular weights were calculated to be 269,000 and 179,000. The DPNH-specific isozyme was synthesized throughout most of the cell cycle of light-dark synchronized cells. The level of the TPHN-specific isozyme appeared to be negligible in the nitrate-cultured cells. The induction of the TPNH-specific isozyme was dependent on both RNA and protein synthesis and the isozyme was inducible at all times during the cell cycle. The maximum rate of induction or the potential increased during the period of DNA synthesis and the fold increase in potential and in DNA were equivalent. This data supports the continuous availability for transcription of the gene for this isozyme in the eucaryote Chlorella and is consistent with the hypothesis that, under fully induced conditions, the gene dosage of the cell governs the potential. / Ph. D.

The Surgical Scheduler's Game is a management game structured within the surgical department of a hospital. It is designed to introduce participants to the scheduling function of the surgical service and is based upon surgical data collected on over 4,000 patients. The participant assumes the position of a surgical scheduler in a 400 bed hospital with five operating rooms. For each operation to be scheduled, the participant receives the name of the surgeon placing the scheduling request, the pre-operative procedure name, the patient's age and the date and time requested. The participant must then schedule the patient. He has available an estimate of the duration of different operations based upon historical records. Other actions required of the participant are to schedule modifications generated through cancellation and rescheduling requests. At the completion of each game period (one day), the participant receives a performance report. This report includes both a measure of scheduling effectiveness and cost of the service that had been performed. In addition, a comparison is made between the best possible schedule for each day and the schedule developed by the participant. The financial report includes fixed, variable and overtime operating room costs, surgeon and patient delay costs, rescheduling costs, and income from operations. The game is set within a probabilistic environment and includes an adaptive learning feature which makes the scheduling requirements more difficulty as the participant reaches certain levels of proficiency. / Ph. D.

This dissertation is concerned with the estimation of three functions of the binomial probability of "success": estimation of a linear combination of independent binomial probabilities; fixed precision estimation of the binomial probability; and estimation of the logit transformation of the binomial probability. A Bayesian viewpoint is adopted temporarily to "suggest" a wide class of admissible estimators for each problem. Designated the class C of SBP estimators, it is the class of Bayes estimators derived from Symmetric Beta Priors (the class of conjugate priors for the binomial), and often includes the maximum likelihood estimator as a special case. Once a class of estimators for each problem is suggested by the Bayesian viewpoint, three criteria are used to obtain the "optimum" estimators in that class. Two of these criteria are classical in nature: minimax risk and minimax weighted risk. The third criterion, the solution of which is the estimator corresponding to the "least favorable" prior in the class of priors considered, is subjective in nature and would appeal more to Bayesians. For each of the three problems, a class C of SBP estimators is suggested, and the optimum estimators from this class are obtained. In addition, for a special case in the estimation of a linear combination of binomials, an estimator is found that is minimax among all estimators, as well as minimax among SBP estimators. / Ph. D.

Chitinous products were investigated by in vivo rumen fermentation. Samples of cockroaches, grasshoppers, shrimp shells, crabmeal, and purified chitin were placed in the rumen of fistulated steers. Average rumen solubility (weight loss from a nylon bag) of 66.5. 32.0, 17.4, 35.7, and 21.5%, respectively, were observed. Measurements of water solubility in a non-rumen system indicated that the rumen system was solubilizing 8, 7, and 12% of cockroaches, grasshoppers, and crabmeal, respectively, over that in water. A feeding trial was conducted where crabmeal was evaluated at 0, 10, and 20% of a basal diet. The experiment was conducted by using six ruminating calves in two replications of a 3 x 3 Latin square design. In each 4-week period, nitrogen balance data was collected during the third week. Crabmeal nutritive value was established. It substituted equally for base ration at the levels studied. No treatment differences were detected for gain, feed intake, absorbed and retained nitrogen, dry matter, or crude fiber digestibility (P<.05). In a second feeding trial using a 2 x 2 Latin square design, a zero and 20% diet were fed to two groups of five ruminating calves for two time periods. Nitrogen balance data collected as in Trial I on two animals of each group, further established crabmeal nutritive value. There was no difference between groups for all performance parameters measured except that animals fed crabmeal had a lowered nitrogen retention (P<.05). Chitin digestion in Trial I averaged 58 ± 19% and 69 ± 22% for 10 and 20% crabmeal diets, respectively; for Trial II, chitin digestion averaged 72 ± 12%. In in vitro incubations of purified chitin and insect skins in strained rumen fluid for three hours, there was no apparent deacetylation of chitin beyond that observed in positive and negative control samples. / Ph. D.

The blue-green alga, Synechococcus lividus, was isolated free from contaminating bacteria by selecting colonies off agar plates which had been maintained in an atmosphere of 4% CO₂-air. Attempts were made to synchronize the algae by intermittent illumination, isopycnic centrifugation, and differential centrifugation. Cells selected by differential centrifugation were the only ones to give acceptable synchrony with cell number increasing in a short division time and a high cross-wall index (85%). Because the density of the green alga Chlorella sorokiniana appears to increase during the cell cycle, synchronous cultures of this organism were successfully selected with isopycnic centrifugation. The enzyme phosphoribosylglycinamide synthetase was found to be stable in Chlorella cells frozen and thawed once, provided substrates were present in the buffer. When frozen-thawed cells or fresh cells were broken in the French press or by a sonic oscillator the enzyme was no longer dependent on substrates for stability. However, cells broken by sonication showed an ATP-MgCl₂ dependent linear loss of enzyme activity on incubation at 38.5°. Breaking fresh cells twice in the French press in the presence of substrates, yielded an enzyme preparation with the highest activity and stability. This enzyme increased in a continuous exponential fashion, paralleling the increase in total protein, during the cell cycle of cells selected by the isopycnic procedure. Because the enzyme continued to increase at the same exponential rate during the period of DNA replication, the increase in gene dosage did not appear to affect the increase in activity of the enzyme during the cell cycle. / Ph. D.

Three parameters, i.e., misorientation angle, misorientation axis, and boundary normal, have been used to describe a general bicrystal, such as two adjacent grains, subgrains, twins, or two neighboring phases. A computer program has been developed to determine these three parameters. The angle and axis of misorientation are calculated from a misorientation matrix which is obtained by using two Kikuchi patterns, one taken from each of the two crystals. To obtain the boundary normal, a specimen tilt inside the microscope is required. A rotation matrix specifying the actual specimen tilt is formulated from two Kikuchi patterns taken from the same crystal before and after tilt. With this rotation matrix and the change of projected boundary images before and after tilt, the boundary normal can be calculated. It has been demonstrated that for high-angle bicrystal: the misorientation may be determined to within ±0.5°, and the misorientation axis to within ±0.2°. For low-angle bicrystals, the misfit angle can be obtained to within ±0.1°, and the axis of misfit to within ±4°. The boundary normals so determined are generally accurate to ±2° if suitable correction is made for magnification changes during crystal tilt. Variations in magnification and camera length due to the shifting of specimen position along the electron-optical axis were investigated. It was found that a variation of 20% in both magnification and camera length may result when a tilting stage is used. A calibration curve was obtained which allows for correction of these errors in the Siemens Elmiskop lA. The inherent accuracy of various beam axis solutions from a Kikuchi pattern, i.e., 3-pole, 3-normal, 2-pole/l-normal, 1-pole/2-normal, and 1-pole/matrix solutions, was also analyzed. The results indicated that the 3-normal solution is the most accurate one. The beam axis thus determined is accurate to ±0.05°, and is nearly independent of the effective camera length. For solutions in which at least one Kikuchi pole is used to formulate the equations, the beam axis may be obtained to ±0.1°, if the effective camera length is calculated from the Kikuchi pole separation. In order to eliminate the tedious calculations required to index Kikuchi patterns, computer programs were developed to provide for computer plotting of standard stereographic Kikuchi projections of any desired orientation and projection sphere radius for FCC, BCC, diamond cubic, and HCP crystals. The programs also provide for identification of Kikuchi poles as well as identification of individual Kikuchi lines. Indexing of observed patterns becomes a matter of comparison between the patterns and the projections. Coates projections for scanning electron microscopy and transmitted Kassel or pseudo-Kassel projections for x-ray diffraction studies can also be generated simply by changing the input wavelength. / Ph. D.

A theoretical analysis is presented for the supersonic flutter of flat rectangular, biaxially stressed sandwich panels. On the basis of two-dimensional static aerodynamic theory an exact solution is obtained for panels having simply supported edges parallel to the airflow. The leading and trailing edges may be simply supported or clamped. The mathematical model describing the panel motion includes terms that account for rotary inertia, face sheet bending deformations, and transverse shear deformations of an orthotropic core such as honeycomb. Damping forces are neglected. The linear system of partial differential equations governing the lateral deflection and the two transverse shear angles is eighth order and has constant coefficients. For simply supported edges parallel to the flow these equations have a separable solution. The characteristic equation associated with this solution is an eighth-order polynomial; its roots are determined numerically for given values of dynamic pressure and panel frequency. For each of the two sets of leading and trailing-edge boundary conditions considered, a corresponding frequency determinant is obtained that provides a second relation between the dynamic pressure and frequency. For either set of boundary conditions the corresponding frequency determinant equation and the characteristic equation are simultaneously satisfied by various combinations of dynamic pressure and frequency. Flutter occurs when the dynamic pressure attains a level that causes two panel frequencies to coalesce and become complex. For a panel with an isotropic core, a second-order differential operator factors from the eighth-order system of differential equations. The remaining sixth-order system has been used in a previous flutter analysis (but with rotary inertia neglected). Flutter solutions based on the sixth-order system are shown to be correct when the complete solution from the eighth-order system uncouples. (This occurs for the isotropic panel if all edges are simply supported.) The sixth-order system, however, is inherently incomplete and is generally not applicable to the more general case where the solutions from the eighth-order system do not uncouple. Numerical results from the complete eighth-order system show that the face sheet bending stiffness has a negligible effect on flutter if the faces are thin compared to the core thickness. As the face-to-core thickness ratio increases, the face bending stiffness becomes more important, especially for panels having relatively flexible cores. The frequency coalescence behavior can be markedly changed by the combined effects of rotary inertia and core shear flexibility. Failure to account for this combined effect can lead to significant overestimates of flutter dynamic pressure values. Rotary inertia also causes the flutter solution to depend slightly on the crossflow in-plane loading. The directional shear stiffness properties of the core are of comparable importance for square panels. As the panel width increases, the importance of the shear stiffness in the crossflow direction decreases. / Ph. D.