The relationships of the fourteen (14) species of Campylobacter were derived by comparison of the partial 16S ribosomal RNA sequences determined using reverse transcriptase and oligonucleotide primers specific for regions in the 16S rRNA molecule. These species formed three distinct RNA sequence homology groups. We propose that the following species remain in the genus Campylobacter (homology group I): C. coli, C. jejuni, C. laridis, C. fetus, C. hyointestinalis, C. concisus, C. mucosalis, C. sputorum and the "catalase-negative to weakly-positive" (CNW) strains. C. pylori, "C. cinaedi" and "C. fennelliae'' were not related to the true campylobacters at the genus level. However, they were related to each other and to Wolinella succinogenes at the genus level and constituted homology group II. These four species should be reclassified and placed within a single genus based on 16S rRNA sequence similarity. C. cryaerophila and C. nitrofigilis also exhibited a high level of RNA sequence homology with each other but not with any other species tested; they constituted homology group III and should be considered as being a single genus. A comparison of the 165 rRNA sequence data from the three homology groups with the same sequences from representatives of the three major phylogenetic branches of the purple bacteria indicated that these genera form a single phylogenetic branch which is only distantly related to the purple bacteria. / Master of Science

Distribution of the orangefin madtom (Noturus gilberti) was determined from 347 sites sampled in Virginia and North Carolina. This species inhabited 264 stream kilometers, over twice the reported range, in the following systems: Craig Creek, Roanoke River, Dan River, Big Chestnut Creek, South Mayo River, Pigg River, and Smith River. The orangefin madtom was somewhat common; 33% (Dan River) to 70% (Craig Creek) of the sites sampled were occupied. Negative interspecific associates of orangefin madtoms included chubs, mountain redbelly dace, rosyside dace, crescent shiners, and crayfish; only Roanoke darters were considered positive associates. Sand and silt levels were significantly lower at sites with N. gilberti, while percentage of small cobble, local gradient, and depth were significantly higher. Discriminant function analysis identified large gravel, local gradient, silt, and occurrence of rosyside dace and crayfish, as significant predictors of the occurrence of the orangefin madtom. Seasonal samples from Craig Creek consisted of three age groups. The smallest individual captured was 33 mm total length (TL) and the largest was 111 mm TL. Mortality appeared moderate from age I until summer of the third year, when most individuals apparently died shortly after spawning. Spawning habitat of orangefin madtoms in Craig Creek appeared to be fast-water riffles dominated by small cobble substrate. Attempts to induce N. gilberti spawning in the laboratory were unsuccessful due to high mortality of captive fish. Field-collected margined madtom egg masses, however, were successfully hatched, and subsequent survival was significantly greater for fry fed ground trout chow versus live brine shrimp nauplii. / Master of Science

Adult female Sprague-Dawley rats, age 10 months, were used to estimate the minimum dietary lysine requirement for tissue maintenance. Ten animals were assigned to one of eight treatment groups by weight. The dietary lysine levels ranged from 0.06 to 0.36 % of diet and the feeding period lasted 56 days. Carcass and liver nitrogen and total serum proteins were determined, and a dietary lysine requirement was estimated from the data obtained. Carcass and liver analysis included weight, total nitrogen, percent protein, percent water and percent fat. Using a one-way analysis of variance, results showed no significant differences in carcass or liver composition between the treatment groups. The data indicated that the mature female rat has a dietary lysine requirement lower than 0.06 % of diet, or less than 20.1 mg/day/kg0.75. Since previous investigators used a protein based diet, a possible cause for the insignificant differences between values seen in this study may be a consequence of using a nonprotein, amino acid mix base supplemented with lysine. In future studies for determining the dietary lysine requirement in the adult female rat, dietary lysine levels below 0.06 % of diet must be included when using a nonprotein based diet. / M.S.

The water distribution system design problem consists of finding a minimum cost combination of network layout and sizes of system components so as to satisfy flow demands, minimum and maximum head requirements and a reliability criterion. A two step procedure is proposed to find a near optimal design. The first step considers obtaining a near optimal tree layout using a heuristic tree search algorithm. Two different methods are followed for the tree search - one for single source networks and the other for multiple source networks. The second step adds loop forming redundant links to the tree layout in such a way that every demand node has two paths from source node(s). The methodology is applied to a single source network and a multiple source network. In both the cases better results arc achieved than those obtained previously by other researchers. / Master of Science

The relationships between molecular parameters and microdomain formation of a variety of block- and graft-copolymers were studied by Transmission Electron Microscopy (TEM). Molecular variables included chemical composition: dimethyl-, fluoropropyl and diphenyl-siloxane , sulfone styrene, paramethylstyrene, t-butylstyrene, arylester and methyl methacrylate, as well as molecular weight and distribution. Effects of the kinetics of phase-separation were also determined . Thick (approximately lmm) films cast from solvent showed more complete phase separation than either thin (about 10nm) cast films or compression-molded specimens. Spherical domains formed in alternating poly(ester/siloxanes), and phase mixing seemed to correlate with the solubility parameters of the three siloxane types. Shear-stresses during molding changed domain shapes and eliminated short-range ordering. In the PMMA-graft-dimethyl siloxane system, SK, 10K and 20K <M_n> siloxanes were incorporated at 16% and 45% by weight. At 16%, spherical siloxane domains formed in both thick- and thin-cast films. The domain sizes and interdomain distances scaled with siloxane molecular weight and total block molecular weight respectively to a 2/3 power law in excellent agreement with theoretical predictions for di- and triblock copolymers. Thin films cast from the 45% siloxane graft copolymers also showed spherical domains with sizes dependent on molecular weight. However, the thick films showed phase transitions from disordered bicontinuous (M_n = 5K) to lamellar (M_n = 10K) to cylindrical <M_n = 20K). Qualitative TEM/EDX analysis of other systems was used to identify oligomers, homopolymers, and contaminants, thus monitoring the effects of novel reaction conditions and work-up procedures. / Master of Science

Several mathematical models are developed for a hose-drogue refueling system in an attempt to represent the physical system accurately and to subsequently observe the dynamic response of the system under different initial conditions. The mathematical models examined include a flexible hose model and a model which includes elastic bending effects. The equations of motion include aerodynamic, gravitational, and tensile forces, and solutions of the refueling system are found using fewer assumptions than in previous work. Once the equations of motion are developed, they are separated into equilibrium and perturbation portions. Solutions of the nonlinear equilibrium tension distribution are obtained by solving the equations in closed form using a two point boundary value problem solver program. The solution to the linear equilibrium tension distribution is found and compared to the nonlinear solutions. Results indicate that the behavior of the solutions is similar, but the linear solution gives larger values of tension near the hose attachment point. The perturbation equation is discretized using a finite difference scheme and the resulting first order differential matrix equation is integrated to calculate the dynamic response for given parameters and initial conditions with the various equilibrium tension distribution solutions. Results show negligible differences between the different tension values upon substitution and it is therefore recommended that the linear approximation to the equilibrium tension distribution be used in analysis of this hose-drogue refueling system because of the ease in obtaining solutions with this method. / M.S.

The first International Coffee Agreement (ICA) came into force on September 1, 1963 and established a system for regulating coffee exports and imports primarily by employing quotas to limit coffee supply. The ICA has been in existence since that time and is now in its fourth edition. The continued operation of the Agreement for nearly twenty five years is an apparent contradiction of economic cartel theory. Two major areas related to the success of the ICA cartelization effort are the focus of this paper. First, the structure of the Agreement, especially the provisions for regulating exports and imports, is analyzed from the perspective of standard economic cartel theory. This analysis highlights several cartel mechanisms incorporated in the ICA. It concludes that the Agreement is soundly structured and allows for the successful cartelization of the coffee market by those countries which voluntarily participate in the ICA. However, this sound structure alone does not explain the success of the coffee cartel because it cannot prevent countries from withdrawing their support of the Agreement. The analysis then turns to the questions surrounding the continued endorsement of the ICA by both coffee producing and consuming countries. Political and economic motivations of the governments which support the Agreement and of the producers and consumers in these countries are discussed. Although some incomplete explanations for their support of the ICA are presented, several questions surrounding the economic reasoning of many parties to the Agreement, especially coffee consuming countries, remain unresolved. / M.A.

The intent of this research is to provide a performance analysis of multiple access protocols in packet-switched Very Small Aperture Terminal (VSAT) satellite communication networks. This research consisted of three major thrusts. First, we analyzed the average time delay of the broadcast Time Division Multiplexing (TDM) outbound channel (hub to VSAT). Second, a throughput performance analysis of an asynchronous Direct-Sequence Code Division Multiple Access (DS-CDMA) communication system is carried out for the inbound line (VSAT to hub). Each channel was characterized by its bit error probability, and transmits fixed-length packets generated according to a Poisson process in an unslotted environment. Third, we presented a delay analysis of the ALOHA DS-CDMA/TDM channel to determine the total service time of a packet originating from either the VSAT or the hub station. In addition to its multiple access capability, this thesis is concerned with the use of direct-sequence spread-spectrum signaling primarily because of its ability to combat interference. Emphasis is placed on average throughput performance, and on the average packet delay after solving for the steady state probability generating function of the station queue size. Then, a discussion of the effect of finite buffer size, and an analysis relating the probability of buffer overflow to packet statistics and buffer size is presented. Because of the bursty nature of a traffic originating from the VSAT’s, the Automatic Repeat Request (ARQ) technique used for error control is the Stop-and-Wait (SW) protocol. It is used as a retransmission strategy in both the Asynchronous Time Division Multiplexing (ASTDM) and the ALOHA DS-CDMA channels. / M.S.

This thesis discusses the problem of decomposing rectilinear regions, with or without holes, into a minimum number of rectangles. There are two different types of decomposition considered here : decomposing a figure into non-overlapping parts, called partitioning, and decomposing a figure into possibly overlapping parts, called covering. A method is outlined and proved for solving the above two problems, and algorithms for the solutions of these problems are presented. The partitioning problem can be solved in time O(n⁵ ²), where n is the number of vertices of the figure, whereas the covering problem is exponential in its time complexity. / M.S.

Stress can have a significant effect on the stability of pillars. To get an accurate picture of pillar stability, information is needed not only on the initial stress magnitude and distribution but also on all subsequent stress changes. Sonic methods have the potential to be among the fastest, most economical and least destructive means of stress measurement. Wave propagation velocity has long been applied to the investigation of the upper mantle and crust stress. Recently, it has also been applied to rock burst prediction and mine site investigations. Another parameter for investigating stress in rocks is the attenuation coefficient. Attenuation has been observed to be more sensitive to stress changes than wave velocity; however, the measurement of attenuation is more difficult than that of wave velocity. In this study, the mechanism of sonic attenuation in rock is reviewed. Both the velocity and the attenuation of sound waves in five different rock types under various stress levels were examined in the laboratory. It was found that the relationship between the velocity ratio and stress and that between the attenuation ratio and stress, for a specific rock type, can be expressed by simplified second order polynomial equations. / Master of Science