The times of arrival from southern sources and the rates and directions of migration are now known for many species of forest trees in New England. This information has made possible the reconstruction of generalized vegetation types for the period 14,000 to 9000 yr. B.P., as has been done recently by R. Davis and G. Jacobson for Vermont, New Hampshire, and Maine. The regional pattern, which has been based mainly on studies of fossil pollen from wind-pollinated trees, includes a diverse and biogeographically interesting flora of bryophytes, herbs, and shrubs that only recently is becoming defined through studies of plant macrofossils. The most thoroughly investigated sites, Tom Swamp (Massachusetts), Columbia Bridge (Vermont), and Upper South Branch Pond (Maine) have produced late-glacial floras with numerous calcicoles, some of which are rare or lacking in the present New England flora. Fossils of arctic-alpine species are also present at these sites, which are all at low elevations. A prominent calcicole element exists in the late-glacial fossil record regardless of whether the extant flora near a site contains such plants. This pattern is evident for both seed plants and bryophytes. For areas with acidic soils, the loss of calcicoles is correlated with the late-Pleistocene arrival of spruce (as documented by increases in macrofossils) at 12,800 yr. B.P. in southern New England, and some 2300 radiocarbon years later in north-central Maine. Leaching, humus or litter accumulation, and other aspects of soil genesis may have been responsible for the loss, but what caused the pattern is poorly understood. As viewed on a continental scale and over thousands of years, the elimination of calcicoles has left fragmented ranges and rare occurrences in eastern North America of boreal or northern plants that otherwise are today widespread to the west.