90 Mr. W. Kemp on the Vitality of Seeds. 



Mr. Darwin) to Professor Lindley^ and sowed the others myself. 

 The plants reared by myself were sent to Professor Henslow, who 

 states that they consist of Polygonum convolvulus and a variety of 

 Atriplex patula ; the seeds planted at the Horticultural Society 

 by the kindness of Professor Lindley produced Rumex acetosella 

 and an Atrijjlex, which was not at hrst recognised, but which 

 Mr. Babington states is exactly like a variety of A. angustifolia 

 which he has seen growing on mud in salt-marshes and on ma- 

 nure-heaps. 



The sand-quarry is situated about a quarter of a mile west of 

 Melrose, and at the height of between hfty and sixty feet above 

 the nearest part of the Tweed. The seeds were mingled ivith some 

 decayed vegetable fibres, and formed a layer resting upon another 

 layer, eight inches in thickness, of tine sandy clay. This latter 

 lay over a mass of gravel, which again rested on a great mound 

 belonging to the boulder formation. This mound, which extends 

 about a mile along the middle of the valley, is about ninety feet 

 in thickness, and I believe was formed by the action of glaciers. 

 It contains enormous angidar blocks of rock, and others smoothed 

 and distinctly scored in lines parallel to their longer axes. The 

 layer of sandy clay, on which the seeds rested, was capped by up- 

 wards of twenty-live feet in thickness of distinctly stratified sand, 

 which has been largely quarried. The beds of sand vary in thick- 

 ness and in fineness ; sometimes they alternate with thin seams 

 of impalpable clay, and sometimes they contain minute pebbles 

 and fragments of carbonaceous, decayed wood. The layers slope 

 at an angle of fifteen degrees towards the valley, and in this di- 

 rection they thin out ; the upper layers extend further into the 

 valley than the lower ones ; the entire mass has a level top, and 

 is capped by some thin beds of fine gravel. From these several 

 facts (as every geologist will admit), and from the general aspect 

 of the layers of sand, it is scarcely possible to doubt that they 

 were deposited by a river or torrent, at the point where it entered 

 a sheet of water. I had long been of opinion that the valley of 

 the Tweed in this part must formerly have been occupied by a 

 lake, at a period when a great trap dyke, 100 yards mde, which 

 crosses the valley four miles lower down at Old Melrose, had not 

 been worn through. By an accurate levelling I have ascertained, 

 that the layers of sand lie just beneath that level which a lake 

 M'ould hold, if the barrier at Old Melrose were reclosed. A de- 

 pression on the surface of the land can, also, be distinctly followed 

 from the spot where the sand-quarry is situated, up the valley, to 

 where it joins the bed of the existing river ; I cannot doubt that 

 the Tweed anciently flowed in this depression, and deposited on 

 the borders of the lake, the layers of sand where we now find them. 

 It is certain that in the time of the Romans, about 2000 years 



