L'objectif de la thèse est de réaliser une mise au point critique portant sur la spécificité des méthodes cartographiques généralement employées pour la représentation quantitative des phénomènes de répartition. Dans un premier temps, cette mise au point permet de dégager, grâce à une analyse de l’information et en tenant compte des règles définies par la sémiologie graphique, les méthodes cartographiques qui, du point de vue méthodologique, sont les plus satisfaisantes, c'est-à-dire, qui correspondent à la nature de l'information à communiquer. Dans un second temps, cette mise au point porte sur la spécificité des diverses méthodes cartographiques à répondre à certains besoins des utilisateurs. Le premier besoin retenu dans l’analyse est la précision de la carte ou sa généralisation. L'utilisation de deux critères, la fidélité d'une part à la nature de 1'information et d'autre part aux caractéristiques spatiales de la distribution de cette information, permet de classifier les diverses méthodes cartographiques sur une échelle précision-généralisation. Le second besoin que l'on introduit dans cette étude est 1'automatisation de la rédaction cartographique. Quelques constatations générales portant sur les contraintes inhérentes à ce nouveau besoin, introduisent une brève critique de divers programmes de cartographie automatique. Enfin, en conclusion, l'on discute brièvement d'un troisième besoin que 1'utilisateur peut manifester, à savoir l'analyse spatiale de la distribution d'un phénomène. L'on montre finalement comment cet utilisateur peut, par le choix de certaines méthodes cartographiques, satisfaire simultanément, quelques-uns de ses besoins, c'est-à-dire, comment il peut intégrer le choix des diverses méthodes cartographiques qui permettent d'aborder les divers aspects tant cartographiques que statistiques de la répartition d'une population.

No description available.

No description available.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

In Virginia, the conventional method of farm production of processing tomatoes has lost its competitiveness to other farm enterprises due mainly to intensive labor requirements and low economic profit. Also, the tomato processing industry in Virginia and other Mid-Atlantic states has lost its market share to other competing regions in the national market due primarily to high raw product cost and relatively inefficient processing operations. Recently, a capital-intensive production method for processing tomatoes, i.e., high-yield varieties adapted to direct seeding and machine harvesting, has been developed in the Mid-Atlantic region, and many tomato growers in Virginia have indicated intentions to adopt the new production alternative. Under such circumstances, the major concern of this study was directed to determine whether or not the adoption of the new tomato production system would make Virginia's processing tomato industry more competitive, not only with other crops at the farm production level, but also with other major tomato producing regions in the national market. Since the basic objective of this study is twofold, two major analytical sections are developed. In the feasibility analysis section, farm production costs of three alternatives - (1) transplant method of conventional varieties with hand harvesting, (2) direct seeding method of new tomato varieties with hand harvesting, and (3) direct seeding method of new tomato varieties with machine harvesting, are compared. In the interregional analysis section, an interregional competition model especially designed for the U. S. processing tomato industry is formulated. In this model, five major processed tomato products, fourteen production states and ten consumption regions were included. The important findings of the analyses were: (1) the average production cost of the mechanical production method was cheaper than conventional production method; (2) the average net return of the new tomato varieties was higher than the net returns from most other farm crops; (3) the quality of machine harvested tomatoes was superior to hand harvested tomatoes for tomato processing operations; (4) the adoption of the mechanical production method would make Virginia's tomato processors directly competitive with California processors in the major Eastern markets; (5) the potential for Midwestern tomato processors to become Virginia's major competitor in the Eastern markets would be strong in the future. / Ph. D.

Significant differences involving near-wall velocity profiles in two-dimensional turbulent boundary layers when such data was plotted on law of the wall coordinates were reported by D. Coles in the Stanford Conference. These discrepancies can be especially important when near-wall velocity data is used to infer the local wall shear stress from law of the wall similarity laws. With emphasis on the near-wall measurements, this study systematically investigated the effects of the following parameters on the velocity profile: probe type (straight-stem claw probe, gooseneck probe, and hot-wire probe), probe tip diameter, streamwise pressure gradient, and boundary layer tripping. In addition, the local wall shear stress was inferred from the near-wall data, and the effects from the previously mentioned parameters and the choice of both the law of the wall constants and formulas on the inferred wall shear stress were examined. / M. S.

A postulate of this research was that useful semiconductors could be made in the iron oxide thin film system by controlling the ratio of Fe²⁺ to Fe³⁺ ions. For the added versatility of producing all of the oxide phases from one source (target), reactive (in an oxygen/argon atmosphere) sputtering was chosen. In addition, to avoid the problems of sputtering oxides by Electron Beam evaporation or Direct Current sputtering, the recently developed technique of Radio Frequency sputtering was employed. Thus, the iron oxide thin films were produced by reactive Radio Frequency sputtering. The correlation between the electrical resistivity and the sputtering atmosphere (oxygen pressure) was made and plotted. Identification of the Fe²⁺ to Fe³⁺ ion ratio was impossible, but X-ray diffraction data were used to identify the iron oxide phase or phases present in each thin film. Thin films which were a mixture of oxide phases were produced and the general sputtering parameters (electrical power settings and oxygen level) required to produce these oxide mixtures have been identified. The deposition time was found to have an optimum of about ten minutes, and times in excess of this produced thin films of equivalent thickness. Differences between phase equilibria predictions and the actual phases formed during sputtering have led to the conclusion that reactive Radio Frequency sputtering of iron oxide is a non-equilibrium process. / M. S.

The method of multiple scales is used to analyze the nonlinear effects of the gas motion as well as the acoustic lining material on the propagation and attenuation of sound in two-dimensional and circular ducts of uniform cross sections in the absence of mean flow. The ducts are lined with a point-reacting acoustic material consisting of a porous sheet or of a perforated plate followed by honey-comb cavities and backed by the impervious walls of the ducts. The effect of the acoustic material is included either by characterizing the material by a semi-empirical nonlinear impedance or by coupling the waves in the duct with those in the liner. Analytical expressions are obtained for the absorption coefficient in terms of the sound frequency, the physical properties of the porous sheet or the perforated plate, and the geometrical parameters of the flow configuration. The results show that the nonlinearity flattens and broadens the absorption vs. sound frequency curve, irrespective of the geometrical dimensions or the porous material acoustic properties, in agreement with the experimental observations. The effect of the gas nonlinearity increases with increasing sound frequency, whereas the effect of the material non- linearity decreases with increasing sound frequency. Moreover, the sharper the resonant tuning is the narrower the adverse frequency bandwidth is. / Ph. D.

This dissertation examines three aspects of banking behavior: (a) involvement of banks in inflationary processes, (b) the effects of inflation on bank earnings and (c) the portfolio behavior of banks in an inflationary period. The study first traces the evolution of the"credit theory of inflation" from the eighteenth century to the early twentieth, and finds that under certain conditions banks may ignite an inflationary process by failing to adjust their loan rates to the real rate of interest. The analysis of the second aspect of banking behavior is carried out at the micro-economic level. After a critical appraisal of Martin Bailey's macro model of banking behavior under conditions of fully anticipated inflation and a competitive market structure, the study finds that Bailey's conclusion as to the favorable effects of inflation on bank profit is inconsistent with his model. The study revises Bailey's model and develops two models of banking behavior along the lines of the neoclassical approach of manufacturing firms. The first model examines banking performance under conditions of a competitive market structure and fully anticipated inflation. The model predicts that under these circumstances the banking system is more likely to lose than gain from inflation. The second model analyzes banking behavior when the banking market is characterized as monopolistic. It is inferred that no definite conclusions may be drawn with regard to the effects of inflation on bank earnings without an empirical knowledge of the parameters of the demand for loans and the supply of funds to the banking system. The study also analyzes banking behavior under conditions of unanticipated inflation. It is found that if savers respond to changes in the rates of interest only after a time lag, or if their subjective probability distribution of expected returns on income earning assets is biased downward, banks are in a position to earn some windfall gains from inflation. The third section of the dissertation analyzes bank portfolio behavior under both anticipated and unanticipated inflation. It is shown that unanticipated inflation forces the banks to alter their portfolio of income earning assets, and it is argued that such activities by banks may nullify the restrictive monetary policies that the monetary authority may impose upon the money market to slow down the rate of inflation. The study also presents evidence on the profitability of Chilean commercial banks during 1937 to 1950, and finds that due to the special rediscounting provisions of the Chilean Central Bank these banks earned more than average rates of return on their capital outlay, The study also suggests that the Chilean commercial banks were instrumental in perpetuating the inflation. Evidence is also presented on the portfolio behavior of U.S. commercial banks for the period 1950 to 1970. It is shown that the U.S. commercial banks moved from long-term assets to short-term assets during all the expansion periods observed between 1950 to 1970. This behavior by U.S. commercial banks is explained by the differential impact of the unanticipated inflation on short-term and long-term rates of interest for the period under review. / Doctor of Philosophy

The effects of contaminated supply air on the performance of a laminar proportional amplifier were experimentally investigated. The air supply was contaminated with oil vapor and particulate matter. Characteristic gain curves were obtained after each stage of contamination for various loading conditions. Photographs showing the location of contaminant deposits were taken. Two inlet geometries were studied: right-angle and straightthrough. The effects of maintaining a constant pressure and constant flow rate at the inlet throughout the duration of the tests were studied. Also, aspect ratio effects. were studied. Rapid deterioration of performance was evident with the right-angle entry due to inlet blockage. By milling a cavity in the bottom cover plate, the detrimental effects of inlet blockage were delayed. With constant pressure inlet conditions and straight-through geometry, significant buildup occurred in the nozzle region and downstream. Decrease in pressure recovery was linear with time. It was determined that null shift was caused by asymmetric buildup in either the nozzle region or splitter region. With constant flow inlet conditions, the damaging effects on performance were delayed for both inlet geometries. Also, for the straight-through inlet, the nozzle region was relatively clean when compared with the constant pressure inlet case. Null shift was found to be the result of asymmetric buildup in the downstream region. At lower aspect ratios, the damaging effects of contamination were more severe and occurred in less time. / Master of Science