70 per cent. These are replaced by first-year young which make up the greater part of the breeding population. The remaining 25 per cent presumably shift their ranges sufficiently in the course of a year to have moved beyond the limits of an area of the size encompassed by the present study.

No precise measurement of the population density on the study area was obtained. It was not practical to capture every individual present there, and rapid population turnover, due to mortality and wandering, obscured the trends. The information obtained concerning movements of opossums suggest that one may habitually forage as much as 900 feet from its home base. Assuming that 900 feet is the typical cruising radius, the areas drawn upon by the trap lines in the three different seasons were approximately as follows: 1949-50—400 acres; 1950-51—350 acres; 1951-52—220 acres. In these same three seasons the numbers of opossums caught were, respectively, 46, 37, and 30. If these figures represent the numbers actually present, densities of one to 8.7 acres, one to 9.5 acres, and one to 7.3 acres are indicated. However, some opossums using the area probably were missed; and on the other hand, not all those caught in the course of a season were present there simultaneously. Many of those present early in the season would have moved away a few months later, and others would have moved in, replacing them. The number present at any one time could scarcely have been more than half the number caught in the entire season.

Crude census-figures were obtained by utilizing the Lincoln Index and computing the total on the basis of the ratio of marked (and recognizable) individuals to others caught in a sampling period. A large number of census figures were obtained over the three-year period of the study. Each separate census, however, was based on an inadequate sample as the number of marked individuals taken at each sampling, as recaptures from the previous sampling period, varied from one to five. While little confidence can be placed in any one census computation, the trends of figures from series of such computations reveal the approximate number of opossums on the area if due allowance is made for certain distorting factors. Presumably the differences in figures obtained at different samplings result chiefly from the margin of error in the data, although it is true that there is rapid change in the actual number of opossums.

The number of active opossums in the region of the study reaches a peak in late summer and early fall, when second litters of young have grown large enough to become independent. At this season the population contains a high proportion of young of the year. During the ensuing months of fall and winter there is a steady decrease in numbers, through various mortality factors, with no replacement until young are born about the first week of March. These young do not become independent until late May or early June, and during the intervening months there is a further reduction of the adults and yearlings, so that the active population reaches its annual low point in late spring. At that time of year most opossums are in poor physical condition.

The area represented by the opossums trapped totaled more than 500 acres, but not more than 400 acres were within the area drawn upon by the trap line at any one time. Usually the area represented at any one time by the trap line was less—100 to 350 acres, with from 25 to 45 traps. Traps were moved from time to time depending on the distribution of opossum sign and food sources, the weather, and the time available for this study. As a result, successive samples are not strictly comparable and a major source of error is introduced into the census computations. Lack of exact correspondence in the area represented by successive samples would result in a disproportionally small number of recaptures, and an erroneously high census computation. While adequate adjustment cannot be made, examination of the data suggests that census figures are too high, by as much as 50 per cent in many instances as a result of this factor, while in some other instances when there was little or no alteration of a trap line from one period to another, the census figure was not affected. In the winter of 1949-50, the area covered was most extensive, from 350 to 400 acres, and the numbers of opossums taken were correspondingly larger. In the 1950-51 season the area involved was approximately 220 acres, and in the 1951-52 season it was a little less than 200 acres. In view of the census figures obtained and the probable errors, it appears that the opossum population in early autumn is about one to 20 acres, and that by late spring it is reduced to not much more than half that number.

Many of the opossums trapped were suffering from injury, disease, or parasite infestation, and some were in critical conditions. A large adult male trapped on April 2, 1952, seemed to be dying from disease. It was much emaciated and the pelage was sparse and ragged, as if the animal had been sick for a long time. The skin had numerous light-colored pustules 1 to 2 mm. in diameter, and these were especially prominent on the ears, lips, and penis. When released, the opossum was too weak to move away. It was excited by movements of the trapper, and stood erect with violent involuntary rocking movements. After a few seconds it gradually slumped to the ground and subsided into quiescence. On the next day no trace of it could be found.

Most of the opossums caught in summer and early fall had eye infections, and all of them were infested with ticks (Dermacentor variabilis). Sometimes ticks were attached in dense clusters of several dozen on the animal's ears and scattered over other parts of the body.

In March and April, 1950, seven adult opossums were found dead in the traps. None of these showed any evidence of disease or injury and they were normal in appearance except that they were thin. It was concluded that death had resulted from exposure and starvation in the traps in these animals already in critical condition as a result of winter food scarcity and frequent fasting. Up to this time the procedure had been to check the trap line only on alternate days and no mortality had resulted, even in the coldest part of the winter. The implication is that by spring, opossums are in a condition so critical that they are unable to withstand exposure or fasting and die whenever weather conditions are unusually severe.

After these losses in the spring of 1950, trap lines were checked daily. However, in October, 1950, further mortality in traps resulted in the loss of three or more opossums. All three of these were rat-sized young of second litters. These young lacked the abundant supply of fat characteristic of larger opossums in fall, and seemingly were unable to withstand exposure to chilly nights. Such susceptibility to cold might result in heavy mortality in retarded second-litter young when cold weather of autumn is unseasonably early or is unusually severe.

Natural enemies of the opossum on the area include the red-tailed hawk, horned owl and coyote. Because of the opossum's nocturnal habits it is rarely exposed to hawk predation. Food habits of the coyote on the area have not yet been investigated. Numerous instances of horned owl predation on opossums have been recorded in the literature. On January 15, 1950, an owl attacked an opossum caught in a live-trap. The trap was found overturned, and a few feet away were entrails and a quantity of opossum hair where the animal was eaten. Low vegetation in the vicinity had many fine down feathers of the owl clinging to it. On December 24, 1950, the carcass of a small adult opossum was found in a pasture near the edge of the woods. The head and tail were intact, but otherwise little more remained than the spinal column, girdles and larger limb bones. White excreta of a large bird beside the carcass indicated predation by a raptor, probably a horned owl.

On a natural area, the University of Kansas Natural History Reservation, in Douglas County, northeastern Kansas, the population of opossums was studied, chiefly by live-trapping, in the fall-winter-spring seasons of 1949-50, 1950-51 and 1951-52. The study area provided a varied habitat of elm-oak-hickory woodland, pastureland, and fallow fields. Opossums use all parts of it, but concentrate their activities in the woodland.

Opossums