444 SHORT PAPERS AND NOTES

latter stages of the Wisconsin ice age. It also seems likely that, as the ice melted, super-glacial till was extensive. This could have formed the substrate for extensive super-glacial vegetation. Ice melting is rapidly destroying this ecosystem (Fig. 3), as it was in those reported by Tarr and Martin6. Ecosystems on Wiscon-sin-age glaciers would also have been de-stroyed. However, organic matter, nitrogen, and other plant nutrients built up in the superglacial ecosystem could make significant contributions to young post-glacial ecosys-tems. For instance, ecosystems of this age below the Mendenhall Glacier have accu-mulated about 2,000 pounds per acre of nitrogen in and on the soil, not including that in the live vegetation3. Superglacial vegetation could have been important in hastening extension of plant ranges after deglaciation. Seed sources for much newly deglaciated land may have existed on nearby ice. Vegetated, stagnant ice bridges could have been important in the spteaa of vegetation and animals %etween islands and between the mainland and islands