804 MK FRANCIS J. LEWIS 



covered with a few inches of water, and a series of lagoons with the higher tufts of 

 vegetation persisting, forming peat and so increasing the height of the hummocks. The 

 life of the hummocks, however, is limited, as the water, wind, and frost undermine the 

 banks, killing the vegetation on the summit and then reducing the hummock to the 

 level of the lagoons. All stages in the growth and decay of the hummocks are re- 

 presented, from those still in vigorous growth, representing the level of the moorland 

 before inundation, to those which are mere clumps of bare peat just rising above the 

 water. The chief plants on the summits of the still growing hummocks are Empetrum 

 nigrum, Eriophorum polystachion, E. vaginatum, Rhacomitrium lanuginosum, Calluna 

 vulgaris, and Erica Tetralix. It is only a question of time when all the hummocks 

 will have been destroyed and a perfectly flat expanse of bare peat formed, more than 

 a quarter of a mile across. Eventually the stream will cut a channel out of the looser 

 peat, and, as this deepens, the flat expanse of peat will become drier and invasion by 

 vegetation will take place. This apparently has been the history of the flat expanses 

 of peat occasionally met with in the Walls area, which are covered with Carex sps., 

 Scirpus c&spitosus, and a few stunted plants of Calluna vulgaris, Erica Tetralix, and 

 Eriophorum vaginatum, although this origin for these areas was not suspected until 

 the earlier stages were seen in Weisdale. 



The lower forest of birch is found all over this area to within 100 feet of the 

 summits of the hills. None of the older strata are present above this altitude, the 

 whole of the peat consisting of the remains of Eriophorum, Scirpus cwspitosus, 

 Sphagnum, and Rhacomitrium, evidently belonging to the latest stage of peat 

 formation. This agrees with the features observed from other parts of Shetland, viz. 

 that the older stratified deposits with the two arctic beds and forest bed were deposited 

 over areas which from their topographical features were suitable for peat formation, and 

 that, at some period after the second arctic bed, conditions peculiarly favourable for peat 

 formation occurred, which resulted in a cap of unstratiried peat over both the older 

 strata and nearly every part of the previously uncovered regions of Shetland. In the 

 forest bed the trunks of birch usually lie about N. 20° to 40° W. 



Both the first and second arctic beds containing Salix herbacea and Betula nana, with 

 the forest bed between them, are well developed along Marrofieldwater Burn, but in some 

 places along Atlascord Burn the Lower Forest is underlaid by extremely hard, dry peat, 

 so hard that it has to be chopped out with a sharp spade. I have not been able to find 

 any determinable plant remains in this layer, but the impressions of long linear leaves 

 apparently belonging to a Monocotyledon are not infrequent. The layer is 2 feet 4 inches 

 thick and rests upon sandstone. The Salix herbacea now growing in small patches on 

 the summits must be regarded as a relic plant dating from the time of the arctic beds. 

 During the first and second arctic stages it occurred generally down to sea-level, but 

 judging from its invariable absence in the forest bed, it must then have been restricted 

 t o small areas on the summits. Considering the small differences in elevation in this 

 part of Shetland (only amounting to 850 feet), the survival of this plant through all 



