816 MR FRANCIS J. LEWTS 



these strata. This is the more desirable as Gunnar Andersson has lately (1909) 

 published a paper entitled "The Climate of Sweden in the Late- Quaternary Period — 

 Facts and Theories" (7), in which he brings forward some evidence to show that the 

 remains of trees are not confined to one level, but occur scattered throughout the peat. 

 The only definite evidence which he brings forward is from two peat-filled basins in 

 Scandinavia of which measurements were made by Holmboe (10) in 1900. From this 

 rather scanty evidence, Andkrsson abandons the conception of the forest beds as 

 definite strata or horizons, and attempts to show that the numerous cases in which he 

 admits well-marked stratification occurs are due to accidental causes, and that only 

 those cases in which tree remains are found throughout the peat, show the true history 

 of events. Of the two sections measured by Holmboe, one at Maridalen, Norway, is 

 242 m. in length, with a maximum depth of 3'9 m.; the other at Tuemyr in Nitedalen, 

 Norway, is 159 m. in length with a maximum depth of 3"6 m. In both the sections, 

 stools, trunks, and twig layers occur from near the surface to the base, but in the 

 Nitedalen section the tree remains seem to be chiefly aggregated about '5 m. below 

 the surface, although a smaller number occur to the base of the peat. 



It will be best to give first an account of the measurements I made in 1909, and 

 then consider the bearing of these facts on the views held by the two antagonistic 

 schools of Blytt-Sernander and Andersson. The three sections of which I made 

 measurements in 1909 are in Shetland, Kirkcudbrightshire, and Inverness-shire. In the 

 first district the measurements relate to the Lower Forest only, as the Upper Forest 

 does not extend beyond the north coast of Scotland. The Inverness-shire sections 

 contain only the Upper Forest, while the section in Kirkcudbrightshire contains both 

 the Upper and the Lower Forest. 



TJie Lower Forest in Shetland. — The relative position of the strata and their flora 

 has already been described. The position of the measured section w T as chosen, not for 

 the regularity of the forest zone at that place — as, before the bank of peat was cut into, 

 the forest was not visible — but for the fact that the general lie of the ground favoured 

 free drainage from the section. Part of the bank containing the forest bed is shown 

 in PI. I. fig. 2. The photograph was taken before the bank was cut back some 

 5 or 6 feet to expose the forest, which was entirely hidden by the slipping down of 

 the loose peat from above as it became disintegrated with frost and sun. The length 

 of the measured portion of the section was 60 feet. It would have been carried 

 much further but for bad weather, it being impossible to make the measurements 

 and plot on squared paper during continuous rain and high wind. The method 

 employed was as follows : — The bank was cut back until a smooth vertical wall was 

 obtained from the surface to the boulder clay. Ordinates were then drawn at a 

 distance of 1 foot from each other on the wall of peat. The vertical distance of every 

 piece of wood in the section above the glacial drift was then measured, and placed on 

 a sheet of squared paper. Each square of the paper represented 6 inches horizontally 

 and I inch vertically. In this way the vertical position of the wood could be plotted 



