The Problem.- Retail trade is larger in some cities of smaller population than in others of a greater population. Wealth, which indicates purchasing power, seems to play some part in directing the pattern of retail trade in different cities. An investigator wishing to find whether the retail market is over- or under-developed in a city or group of cities may find that, although the population is slightly smaller than in a similar city, indications of wealth are sources of a potential purchasing power greater than that of a city with a larger population. He must consult authorities who have studied similar problems scientifically or work out some method of determining which city to select as a potential market for development. He will find some statements by marketing experts concerning the part played by population and wealth, but in general he will find that available works contain vague, conflicting, or indefinite statements concerning the tendency of retail trade to follow population or wealth without many scientific attempts being made to evaluate these factors and prove these statements. The present study, working on the hypothesis that of the two, population and wealth (purchasing power), one exerts a greater influence on retail trade than the other. To test this hypothesis, this investigation makes use of correlation techniques. Specifically, since the value of a study of this nature would be lessened by attempting to generalize about all parts of the United States, we shall select southeastern cities which are similar to those in Georgia. Next, we shall select a trade which is not affected to a great extent by purchasers from outside the limits of the city. Finally, we shall classify all factors under the headings of trade, population, and purchasing power. Importance of the Problem.- If the analysis shows that either population or wealth is relatively more important than the other, it should be possible to make some generalizations about trade in urban Georgia and in similar southeastern cities. This study may add to the slowly accumulating knowledge of markets which is assuming more and more the aspects of an accurate science. The business census of 1930 first gave the United States and its research workers something more definite than the guesses and estimates with which they were forced to work in former years. Method of Procedure and Data.- The factors relating to population and weal.th are selected by logical methods. We shall make use of factors which have been found to be associated with trade by other investigators, and shall employ the data which are available for 1929. The year 1929 is selected because it is the only year for which accurate data are available on population, retail trade, and income factors. The censuses of business taken in 1933 and 1935 are forced to estimate population, since an accurate count of population is taken only every ten years. The cities, representative of Georgia urban areas, are selected by arranging all states in the Southeast, as defined by Odum a.nd Moore, 1 according to rank in income per capita in non-farm centers. The states selected are Tennessee, Mississippi, Georgia, Florida, and Virginia. The cities are above 30,000 in population. The information on cities with smaller population is too meager to allow the factors to be classified and studied. To measure wealth, we construct an index of purchasing power composed of the most potent factors relating to income properly weighted. This index will be correlated with the population data, estimates of relative importance arrived at, analyzed, classified, and used as the basis for the conclusions of this study. All available models of such a study are analyzed, and their conclusions are compared with conclusions reached in this study. There are limitations to such a study. Many complex factors are hard to measure, such as location, transportation facilities, and competition, which are not taken into account. This is partly overcome by selecting cities which are in the Southeast and which are similar to cities in Georgia in respect of per capita personal income. Selecting a trade not affected by external trade to a great extent should overcome the difficulty caused by rural trade flowing into cities for shopping goods. The data consist of statistics gathered by governmental and business agencies1 for the period of 1929. Summary of Work by Other Investigators.- There is a. relative scarcity of work on problems of this type for at least two reasons: (1) the methods of multiple correlation are relatively new, and only recently scatter diagrams and other graphic devices have been used to reduce the tremendous amount of mathematical work involved; (2) data on business end economic features of the United States have only become available since the first census of retail distribution was taken in 1929 by the United States Bureau of the Census in that year and in 1933 and 1935. For this reason it is difficult to secure data which could be used to construct an index of purchasing power. A number of studies have been made of the effect of out-of-town trade on urban trade and location of stores within cities, but few of these studies have made use of multiple correlation or purchasing power indexes. The following summary illustrates the difficulties encountered in trying to find a general agreement among authorities on the relative influence of wealth and population. Converse and Mitchell feel that larger towns attract more trade than smaller towns. Dr. Margaret G. Reid states that retail activity and centers increase with the population of the community. The U.S. Bureau of the Census stresses the influence of population on wholesale trade and finds a correlation of .84 with population by the rank data method. J. M. Cassels finds that retail trade is concentrated where population is most dense and where consumers have less opportunity to be self-supporting. Inez K. Rolph concludes that population is the more important factor influencing intra-city trade. 5 This group of authorities stresses population. The following group emphasizes wealth or purchasing power. Lawrence B. Mann finds income and banking resources are more important than population. Dr. Ennna Winslow emphasizes the importance of purchasing power in the study of consumer markets and recommends the construction of a statistical index of purchasing power. Riggleman and Frisbee recognize population's importance but lean heavily toward purchasing power as the vital factor in a market. Eaton Van Wert Read stresses purchasing power and likens shopping dollars to magnetized particles drawn more by shopping goods and less by convenience goods. He employs simple, partial, and multiple correlation. John A. Pfanner, Jr. minimizes population and uses multiple and simple correlation as well as scatter diagrams to test twenty four variables connected with wealth.