Hypotheses on the nature and processes of culture can be framed from rough observation and general knowledge of cultural situations, or even from logic and imaginary situations, but to test these assumptions one must have more than theory. Thus, although the archaeologist's raw data are for the most part artifacts, from which his deductions about ancient culture are largely inferred and his discipline's role in contributing to culture theory is correspondingly limited, nevertheless this limitation has its advantage in that it forces this particular branch of anthropology to conclusions via data that, for the very reason that they are objects, can usually be dealt with quantitatively. Linguistics, physical anthropology, and certain aspects of ethnology share this characteristic with archaeology, but only linguistics, genetics, anthropometry, and a few of the ethnological studies (for example, kinship) have taken full advantage of it.

Before I could consider the foregoing processes in various aspects of prehistoric Georgia culture I first had to determine what changes, if any, took place, and in what categories. Also, I had to decide, when it was possible to do' so, to which of my categories of cultural traits each ceramic feature belonged. For example, is fabricmarking a decoration or is it simply a technical product, resulting from wrapping a vessel in a fabric before it was wholly dry? Was surface color of unpainted vessels deliberately controlled by the ancient Indian potters, and thus an aesthetic trait, or were only technical manufacturing motives and accidents of firing involved?

The reader can more easily follow my descriptions of vessel shape if he will bear in mind the following. I describe a segment of vessel profile according to (a) curving quality and (b) angle or flaring quality. Under (a) a profile may be outcurved, rounded, straight or incurved; under (b) it may be flaring, vertical, or insloping. A vessel wall or any part of it may, for example, be at once slightly rounded and vertical, or slightly rounded and flaring, or straight and insloping.

Before presenting the projectile point classifications, I feel that a short summary of the unusual circumstances attending the timing of this report is necessary. I submitted what I considered the final version of the manuscript to the Memoirs editor in March, 1963. All figures were submitted to the engraver at about the same time. The classifying of the projectile points and other lithic artifacts, the descriptive text, and the comparative research had actually been completed almost two years before that, the subsequent period from 1961 to 1963 being devoted, in what spare time I could find for the purpose, to drawing figures 45-146, preparing the photographic layouts of figures 235-252, and assembling the final typed manuscript for publication.

I do not know how one can distinguish with certainty between knives and scrapers, except when the latter are so blunted that they would not readily cut. I suspect that we tend to call an artifact a knife when it is roughly long and narrow, or when its business edge runs lengthwise, or when its shape suggests that it was hafted. Any of these is inadequate basis for identification; scrapers could be hafted, and knives were not necessarily so. I would suppose that any tool that cuts like a knife could also be used as a scraper, although not always vice versa.

For purposes of this study, the Appalachian archaeological region of Georgia will designate that part of the state north of the following course from the South Carolina border in the northeast, southwestward to Alabama: the confluence of the Chatooga and Tallulah Rivers, thence along the northwest boundary of Banks County to a point just south of Gainesville, along an irregular line midway between the Chattahoochee Drainage on the south and the Etowah Drainage on the north. This meandering course passes through northern Forsyth County, south along the right bank of Vickery Creek to Alpharetta, westward through Cobb County north of Marietta to Dallas in Paulding County, thence almost due west, skirting the headwaters of Pumpkinvine Creek and following the Polk-Caralson county line to Alabama