This paper was derived from study, for six years, of moose ( Alces alces ) in Wells Gray Park, British Columbia. Part of this work involved study of the annual migrations of these animals between lowland burns with deciduous vegetation and coniferous forests with few burns among them at higher elevations. A number of mammals are known to be migratory. Hamilton (1939) notes that whales and seals of several species, some bats, caribou, bison, elk, mule deer, mountain sheep and porcupine undertake migrations. The last four move mainly in relation to altitude. Hatter (1946) has noted that moose are migratory in parts of British Columbia, while Hosley (1949) quotes Palmer's reports, on moose in Alaska, which describe seasonal movements similar in many respects to those in Wells Gray Park. Moose ascend the Clearwater Valley in spring when snow is still present on the valley floor. Data on spring migration is obtained from counting moose tracks crossing a snowshoe trail. Tracks are counted daily, are recorded as to direction, then are marked to avoid inclusion in subsequent counts. This technique has been used in California and Oregon for deer (Interstate Deer Herd Committee, 1954). The fall migration is not so easily observed. Its progress on the valley floor is more leisurely, with movement continuing downward throughout the winter. Some information concerning this movement is available from hunting statistics gathered in autumn, while track counts and general observations provide further data. The main objective of these track counts in spring is a partial census of the herd. The first movement north, up the valley, is rather sudden and involves over a thousand animals. Since this first movement is quite variable as to calender date, means for predicting the time of migration are necessary for efficient operation of the track counting project. This paper is …