Keilhack, basing his study on the silting and dune-formation at Swinepforte, estimates that the time elapsed since the maximum of the Littorina depression down to the present has been about 7,000 years, making the date of the depression about 5000 B. C. He considers his estimate as somewhat more probable than De Geer's. Anderson has called attention to the change of position of the earth's axis at different times. When the position of the earth's axis was such as to give most sunlight in Sweden, the midnight sun was above the horizon at Karesuanda, the most northern astronomical station, 62 days. During the time of most unfavorable position it was above the horizon only 38 days, a difference of 24 days. This change should influence climate and vegetation. The period of maximum sunshine, according to this view, was 9,000 years ago, about 7000 B. C., somewhat earlier than the maximum of the Littorina depression. It would tend to give a climatic optimum at nearly the same time as estimated by Menzel. Steenstrup[129] discovered the succession of forest growths in the peat-bogs or moors of Zealand, north of Copenhagen. In the layers of some of the depressions he found what seemed to be almost a complete record of forest life from the time of the retreat of the glaciers. The upper layers of peat contained remains of trees still flourishing in the surrounding country: alders, birches, and beeches. Then came oaks, and still deeper the pines. Beneath these were aspens, arctic willows, and other plants of the far north. Remains of the reindeer occur in their lowest layer. The pines hardly, if at all, reached Denmark before the Ancylus Epoch, preceding periods showing only the Dryas flora. The pines had a hard struggle for life at first. They are dwarfed and their rings of annual growth are very thin, sometimes as many as seventy to the inch of thickness. Still some of these dwarfs attain the very respectable age of 300 to 400 years. Gradually they prospered, and in the upper layers there are trunks more than a metre in diameter. All these facts point to early and long occupation. Steenstrup reckoned the age of the oldest layers of these accumulations at 10,000 to 12,000 years, dating their beginnings therefore at 8000 to 10,000 B. C. Pine was still growing in the neighborhood of the shell-heaps, or the capercailzie or pine partridge would probably not have occurred. But in the shell-heaps we find only oak charcoal, not pine. This was at least beginning to retreat and give place to the oak. At Maglemose we find pine charcoal but oak pollen grains in layers apparently of the same age as the settlement. Placing the shell-heaps in the early part of the pine epoch would date them as early as 7000 B. C., or even earlier, according to this chronometer. Hence the older writers, who placed the shell-heaps in the pine epoch, dated them considerably farther back than we do now.