Make your own free website on
Mound Selection

The most frequent criticism of the mounds study is how the mounds were initially selected. It has been argued that perhaps the number of objects that could be classified as mounds is so great that virtually any desired ordering could be found by arbitrarily including certain mounds and excluding others (Flemming 1998). Mound S is really four mounds, those of the City Square. Crater has come under fire for using the geometric center of these four as the location of “one” mound.
Crater and McDaniel argue that these criticisms are unfounded for two reasons:
  1. The population of mounds would have to be substantially increased in number before the observed ordering could be reduced to a reasonable level of chance.
  2. The selection criteria used to determine the test population was sufficiently stringent. The selection was not arbitrary.
Crater specified the following criteria for mound selection:
  1. Size – area in pixels between 49 and 132.
  2. Albedo – brightness significantly greater than the background.
  3. Shape – faceted or dome-like, most having shadows tapering to a point.
  4. Isolation – a clear separation between the object and other land forms.
With a cursive glance at a blow-up of the City, it can be seen that there aren’t really that many mounds and the selected ones look similar and are generally unique to Cydonia. Crater’s definition of a mound was based primarily on the appearance of mounds A, E, and D because they were the first to be noticed as marking the apexes of an isosceles triangle. These are the archetypes of all the other objects designated as mounds in Crater’s extended set and thus served as the model from which all the others were selected (Flemming1998). It is astonishing that the geometry found among the mounds are based on vertex locations that are very accurately represented by the top-dead-center points on the mounds, not just anywhere within the area covered by that mound.
Archeologists use many statistical methods to calculate the probability that any given set of data scattered across the landscape is random. A popular test used in Europe is called the Kolgomorov-Smirnov test for randomness. When applied to the mounds, this test supports artificiality. “…we are almost at the 1% confidence level (we are 99% sure) in finding that the mound data is not normally distributed, as well as uniformly distributed…from the viewpoint of a practicing archeologist, these calculations indicate that we should take a closer look.” (McDaniel/Strange 1998).

Home Links About Us