skip to content

Darwin Correspondence Project

From William Kemp   2 October 1843


October 2d—1843.—


At your request, I forward the following account of the discovery, and description of the place where that ancient seed was found, a sample of which I sent you last March, and which you inform me you intrusted to Professor Lindley with a request to have it planted in the London Horticultural Garden.1 I am happy to hear that a part of it has germinated, and produced plants, which is now being examined by some of our most celebrated Botanists.2 I have likewise been successful in raising a few plants; specimens of which I have forwarded to Professor Henslow. I planted a part in Pots which I kept in the house, likewise a part in my garden, In both a few seeds seemed to germinate vigorously, however with the greatest care I could scarcely keep those in the house alive, as they continued to be so very delicate, the air in the house seemingly was too dry for them although I admitted it daily by the window. Those in the gardin throve much better: still they seemed to suffer from the same cause, though nothing like to the same extent. From that seeming delicasy I intertained the opinion, that the Climate had been much more moist when the seed was deposited, as most likely that would be the case from the numerous undrained lakes and swamps that would so plentifully overspreat this “land of the mountain and the flood” My curiosity being strongly excited by the discovery, of those seeds, I lost no time in visiting the place, and making minute enquery into all the circumstances. The sand Pit in which they were found is situate upon rising ground, about a quarter of a mile west of the Town of Melrose,3 between the modern parish church and an eminence called the Quarry-hill. For many years the surrounding locality has been supplied with sand from that place. Indeed a considerable hollow has been formed by the yearly removal of that material; the place having formerly been dug over, but the sand being of great depth, they have returned upon it, and are now clearing it out by the bottom upwards of twelve feet below the former excavation. The bottom consists of gravel, overspread with some thin layirs of fine sandy clay. It was upon those clayey beds that the seed referred to was found, below an accumulation of sand, which orignaly must at least have been upwards of 25 feet deep. The seeds are of two kinds, The one which was the most abundant is about the size of turnip seed but slightly flattened, and of a fine glossie black. The other is somewhat triangular, pointed at each end, and about the eight of an inch long. They had evidently been deposited along with plants on which they grew, as a portion of decayed vegitable fibre was found along with them.—

Having from infancy been familiar with the locality where those seeds were found, I was long since led to believe that the valley of the Tweed about Melrose must anciently have been an extensive lake. There is a trap Dyke 100 yards broad which runs across the river about four miles further down, at a place called Old-Melrose, which at one time, has doubtless formed an immense weer, damming back the river for many miles, and could only be cut through in the course of ages, by the action of the water. The discovery of those interesting seeds seemed to throw some light on this hypothisis, and induced me to investigate the place anew. By a series of accurate levellings, I found that the sand bank where the seeds were discovered was lower than the top of the ancient barrier accross the Tweed, and consequently than the surface of the supposed lake. This was so far satisfactory; but the next question which naturally suggested itself was, How came that immense bank of sand to be thrown up in that particular place? It is evidently river sand, but the place is far south from the river, and besides high ground intervenes, trending considerably to the east as a promontory between. By a careful and minute survey of the locality, however, I was speedily led to the conclusion that the sand had been accumulated upon the very edge of what had formerly been the very mouth of the river, where it emptied itself into the lake, and a continuous hollow track, which must have been its bed, running westward upwards of a mile round by the east side of the village of Darnick, and then coinciding with its present bed, is still distinctly traceable. The highest part of that ancient water course is not above forty feet above the present channel of the Tweed at the point of their juncture, and as the lake must have stood above Ten feet higher, the river in the time of flood, would bear along the comminuted sandy particles into the still waters near its mouth, while the finer clayey sediment would float farther into the lake, subsiding at a greater distance; and this is found to have been the case, as the bottom of the valley is of a deep clay, The Tweed having subsequently run out a more direct course towards the north.—

The sand pit lies over the side of a gentle ascending gravel bank, the sand is distinctly stratafied, the beds dipping to the valley at an angle of fifteen degrees. Those beds do not present a regular uniformity, either in thickness or quality, a few inches being fine where not one particle exceeds another in size, than an inch or two much grosser, with occasional pebbles as large as a pea, and here and there a thin seam of impalpable clay. Interstratified with these I frequently find numerous embedded fragments of decayed black vegitable matter. No person upon examining that spot can intertain any other Idea, than that the sand has been deposited through the course of ages by successive floods. The material in the varying stratification evinces that. Besides I observe that the beds thin off towards the top, and are thickest half way down the bank, and the upper beds always running farther into the valley, while over the nearly level top lies a coating of fine gravel, which has been lodged there as the edge of the lake became silted up. It is a curious fact, that the black vegitable substance above aluded to, although evidently wood, is never found in pieces of large size, but always in small angular fragments, from the bulk of Mustard seed to a small bean, with the corners rounded off by attrition. It may seem enigmatical to assert that wood has been reduced into such fragments as represented, but such is the fact, and of no difficult solution. Fresh wood would not subside, neither would it be broken in the manner described, but continue to float onwards until stranded upon the banks; but that had been old decayed wood, which had been swept from mossy ground by the banks of the river, and broken into fragments in its passage, and subsiding by its superior gravity in the still water of the lake.

One of the plants produced by those seeds being a runner, it has doubtless taken hold of the other, most likely they had grown upon a bank by the river, which had been carried away in the time of flood, and locked in close embrace they had been swept along, while a portion of the earth adhearing to their roots had caused them to sink in the lake, and the floods of centuries had deposited that vast accumulation of sand above them.—

How long it may be asked, must those seeds we have referred to have been preserved in the secret repositories of nature? But alace! to this we can give no definite reply. We are assured of this, however, that no lake was there about two thousand years ago when the Romans visited the place. Before that era the Tweed had cut its way through the Trapean barrier, besides running out a ravine more than a mile in length behind that weir Considering all the circumstances of the case, I deemed the discovery so extraordinary that I have been most anxious to ascertain if their could be no deception in the matter, and so thoroughly have I been convinced of its truth, that whatever argument may be adduced against it, my opinion must remain unshaken.—

I am | Sir | your most Obedient | Servant, | William Kemp Charles Darwin Esqr FRS, &c

CD annotations

1.5 germinated] pencil cross in margin
heavily scored pencil
scored ink and ink cross in margin
2.3 100 yards,] underl ink possibly by CD
scored pencil and pencil cross in margin
scored pencil and pencil cross in margin [diagram here] Gattonside hills Gattonside Tweed River Church Darnick Abbey Newstead River Tweed Deep gorge Melrose Old Melrose Sand pit Old Water course Trap dyke Red Felspar Porphyry Quarry hill Eildon hills Trap Dyke This line marks the boundary of the ancient lake Trap Dyke The dotted line, the ancient course of the Tweed Dryburgh Abbey Mile The Sand Pit


Describes germination of seeds found in sand-pit and the geological setting of the sand-pit.

Postulates existence of ancient lake along the course of the River Tweed. Notes evidence of glaciation.

Letter details

Letter no.
William Kemp
Charles Robert Darwin
Sent from
Source of text
DAR 50: A8–11
Physical description
7pp † sketch

Please cite as

Darwin Correspondence Project, “Letter no. 699,” accessed on 15 July 2020,

Also published in The Correspondence of Charles Darwin, vol. 2