In view of these traits, and the difficulty of obtaining a representative sample, no precise measurements of population density can be made. During the time required to secure a sample, the population undergoes change. At the pond rock pile, an area of approximately .05 acre, the skinks were found in remarkably high concentrations, 57 in 1949, 85 in 1950, 37 in 1951, and 51 in 1952. These numbers represent population densities of, respectively, 1120 per acre, 1960 per acre, 746 per acre, and 1000 per acre. No such concentrations were found elsewhere, and probably do not occur in natural habitat. The Skink Woods study area of 2-1/4 acres is typical of favorable habitat in the region of the study, and the numbers taken there are more significant. For 1949 the 74 skinks recorded comprise an incomplete sample, and the population density of 33 per acre represented is certainly somewhat too low. For other years the following population densities (exclusive of hatchlings) are indicated: 1950, 92 per acre; 1951, 61 per acre; 1952, 49 per acre. These figures are only approximate, of course, and it is difficult to judge how accurately they reflect the true numbers. Even the most intensive collecting may be insufficient to obtain every individual on a small area. Within each season there are shifts of range by some individuals, off the study area and corresponding shifts onto it by others, so that the numbers caught in the course of an entire season are somewhat too high. The individuals taken on the study area may regularly range beyond its boundaries to some extent, so that the seeming population density is somewhat too high. Actually this was probably a minor source of error for the Skink Woods study area, as nearly half its perimeter was bordered by an open field uninhabitable for the skinks, and the remaining perimeter adjoined areas much less favorable than the central portion. Census of the population of the study area by a ratio such as the "Lincoln Index" used in game management studies was scarcely practicable because of the changing seasonal habits distorting the recorded ratios of the sexes and of age groups somewhat differently at different stages of the season. These changing ratios tend to produce an erroneously high population figure, unless separate computations are made from the data for adult males, adult females, and young. Census figures obtained by this method were erratic but seemed to bear out in a general way, the population figures based on total numbers of individuals taken. In favorable habitat where they occur in high populations of 50 to 100 per acre in spring, these lizards must attain a biomass of a pound or more per acre. Biomass in a population probably fluctuates but little during the course of the annual cycle, even though the number of individuals changes greatly. The steady elimination of individuals through various mortality factors, is compensated for by rapid growth of the young. Summary