scarce aboveground in January as compared with the younger animals, those probably less than one year old. Further, the proportions of old to young increases rapidly towards the last of that month; in other words, as the breeding season approaches. It is probable that the full old-adult population is not abroad aboveground until the last of February, when the ratio of old adults to young of the previous year would certainly not be nearly so little as 1 to 4, which is the minimum possible at the immediate close of the breeding season.
In spite of the above lines of evidence, the real extent of this habit of æstivation among our ground squirrels is not satisfactorily known. It is exceedingly difficult to follow any individual squirrel under perfectly normal conditions through all its various activities for any great length of time. However, an important factor concerned in the work of destroying these animals is suggested; that is, the desirability of placing emphasis upon the need of poisoning in the spring rather than in the fall, when part of the breeding stock may be stowed away out of the reach of poisoned grain. It is a question, too, whether or not a dormant animal, in which respiration is extremely slow, would be fatally injured by a fumigant before the latter would be dissipated.
Human interest in the California Ground Squirrel naturally concerns itself most especially with the questions of total population, rate of increase, and rate of re-invasion of territory previously cleaned of squirrels. As to the first question, we have found it difficult to find an accurate basis for determining the squirrel population living on any given unit of area, such as an acre or a square mile. Counts may be taken of living squirrels that happen to be aboveground at any one time, or of burrows which give evidence of current use. In the first case the count is never likely to cover all of the squirrels in the area, because the chances are overwhelmingly against all of the squirrels being aboveground at one time. Season of the year, time of day, and state of weather will affect profoundly this proportion of squirrels below ground to those in sight. In the latter case some sort of estimate of ratio of squirrels to burrows must have been arrived at. Season of the year must again figure importantly in the estimate, because of the jump in population following the breeding season, and progressive decrease thereafter.
On July 26, 1918, the authors took two censuses of ground squirrels on a badly infested ranch about three miles northwest of Walnut Creek, Contra Costa County (see fig. 16). The first census was taken on a south-facing hillside on an area 100 feet square, approximately one-fourth of an acre. Three counts gave 19, 16 and 17 squirrels, respectively, in sight at once. Twenty-five open burrows were counted in this area. This, therefore, was at the rate of 76 squirrels and 100 burrows per acre, or 1⅓ burrows to each squirrel. The breeding season at that date was well passed. Allowing one adult to every four young gave fifteen adults to every 100 burrows or between six and seven open burrows to each adult squirrel after the breeding season.
The second census was taken in a mowed field from which a crop of barley hay had been recently harvested. The area taken was on a north slope and measured 250 feet square, covering about 1⅓ acre. Three counts were taken between noon and one o'clock. The number of squirrels seen out at once was 25, 26 and 25, respectively. Sixty burrows
40
