Home Up


This may be 'How Not To' do a scientific experiment?


THE general interest that has been shown in the theory of dew-ponds which we advanced in the first edition induces us to give the results of certain experiments which we have carried out since its publication.

We held, and we hold, that experiments conducted on a laboratory scale would not be conclusive, because, except on a very still night, the air passing over a small chilled surface would not remain in contact with it for a sufficient time to become itself chilled. Further­more, even if chilled, it would not deposit its moisture in the prepared receptacle. We, therefore, determined to construct a large dew-pond, in a thoroughly scientific manner. Had we been able to carry out our original design we believe that it would have been successful even without the further improvements which ex­perience has shown us might have been introduced.

At first we endeavoured to obtain a patch of land at some fair elevation, but, as this was not practicable in any convenient locality, we finally made up our minds to be satisfied with a low-lying site. We trusted to our methods in the construction of the dew-pond to overcome this disadvantage, and considered that if it succeeded in such a situation, the general applicability of the system would have been more conclusively demonstrated.

For the reasons already given, we determined that our pond should have a large superficial area. We began by excavating the ground over a space 100 feet square, thus obtaining a superficial area of 10,000 feet. The excavation was carried to a uniform depth of 1 ft. 6 in., and a layer of 4 in. of concrete was laid over the whole. Upon this we put a coating of pitch to stop any moisture from below from penetrating through the concrete to the layer of non-conducting material that we intended to lay. In order to form a suitable bed for the non-conductor, we spread dry sand over the tarred surface.

After much consideration we selected mica, as used for boiler-covering, for our non-conducting material. This was specially manufactured for us in blocks resembling paving-stones, 2 ft. square, and 2 in. thick. Our original intention was that these blocks of mica should be made with a very thin, impermeable, white enamel, in order to reduce to a minimum any heat-retaining material on their surface, and we intended to run asphalt only into the joints between them.

The manufacturers carried out experiments, but were not able to obtain such a coating for our blocks, and in consequence we were, most unfortunately, compelled to lay a coating of asphalt over the whole upper surface as well as between the edges. We were aware that this was an unscientific departure, but we were confronted by a difficult and disappointing situation. Still we did not abandon the work, for our faith in our non-conducting layer of mica was so strong that we believed it would be sufficient to overcome the disadvantage of the asphalt covering.

Our pond was in process of construction in the early spring of 1906. After the mica blocks had been laid and covered with asphalt, we noticed in the early mornings that pools of water collected in any accidental depression in the surface of the asphalt, though not a drop of rain fell. The efficiency of the system was almost startlingly shown when hoar-frost was formed. At such times our pond looked, as one of the workmen graphically described it, " like a great window lying out on the ground." The surface covered by the black asphalt was then divided into perfectly white squares, nearly two feet across, with hard black lines, about two inches wide, dividing them. This was distinctly encouraging, for it showed that the frozen dew had been deposited upon the surface of the asphalt immediately over each block of mica, and that it was entirely absent over the joints between the blocks, where no non-conductor lay below the asphalt covering. When the hoar-frost melted in the sun's rays later in the day, we found to our satisfaction that the pond contained some hundreds of gallons of water. However, as the day wore on, this completely evaporated. Night by night we used to get our water, and day by day we lost it again.


Thus, in spite of its disadvantages of position and material, our pond was successful up to a certain point, and had we constructed it in the form of a funnel with a draw-off cock at its inverted apex we should have been able to collect our accumulated dew of the night into an underground tank, where it would have been preserved from evaporation during the day.

The layer of asphalt with which the bottom of the pond was covered was half an inch to three-quarters of an inch thick. Asphalt is a splendid heat-retaining substance, and if it once becomes warmed by the heat of the sun, this heat cannot, under the conditions of our experiment, be transmitted to the earth, owing to the non-conducting property of the layer of mica placed immediately below it. Thus it dawned upon us at the time that, if a first-class evaporating-basin were required, we could recommend no better method of constructing it than that adopted in our experimental dew-pond.

Now, however (spring 1907), we are seriously modifying that opinion, for the pond remains practically full in spite of the fact that there is no chance of any surface water entering it. The rain which fell into it during the winter, chiefly in October, is not disappearing, and we shall watch it during the summer with great curiosity.'

1 We find (August 1907) that the British Boy has invaded the experi­mental dew-pond in spite of all our efforts to protect it with barbed wire. In a spirit of pure destructiveness, he has pierced or torn up the asphalt in several places. Thus the water has been admitted to the layer of mica, and reduced it to a useless pulp.


While we were thus occupied, Mrs. George Hubbard constructed a dew-pond on the same principle, but employed different materials, and the results that she obtained were encouraging. Her pond was 24 ft. square, with sloping sides, and its depth in the centre was 6 ft. After the ground had been excavated, it was covered with bituminous sheeting. A layer of wool lined the upper surface of this sheeting, and over the wool was another layer of bituminous sheeting supported by a skeleton of wood arranged to prevent compression of the wool by the weight of water.

In the early mornings slight runnels of water could be seen trickling down the bituminous sheeting, but the pond failed to fill. It was therefore demolished, and it was then found that vegetation had grown through the lower layer of sheeting, and that the wool was sodden with water. Its non-conducting property was therefore destroyed.

The situation of this pond was not a favourable one, and it had the further disadvantage of an over­hanging tree. It was noticeable that this tree interfered with the process of radiation from the pond, and that the side thus overshadowed failed to collect dew.

We do not attempt to give an account of all our experiments, but the results we obtained in connection with the deposition of dew on variously coloured surfaces may have some interest. The greenness of nearly the whole vegetable world led us to consider the possible influence of colour. Roughly speaking, all vegetation is green, and especially so at that period when it is most in need of moisture. The truly parasitical plants, not relying on moisture either from the air or the ground for their sustenance, are not green ; we therefore thought that a green surface possibly had some peculiar power of radiation.

We had four pans made of the same wood, 2 ft. square, painted black, white, red and green respectively, and varnished. If we placed a pan of any colour on one of our blocks of mica, we succeeded in collecting in it more dew than in any of the other pans which were placed upon a gravel path. White, on the whole, gave the best results, though green ran it very closely ; black always gave the worst, and red was not much better. On putting the white pan upon a block of mica, and comparing the amount of dew collected in it with that collected by the black pan unprotected by mica, we found that the white pan collected five times as much as the black. Of course the pans were exposed on the same night, and lay within a few feet of one another.

The object of these experiments was to determine the proper colour-surface for our dew-pond, and, as a result, it was ultimately painted white over the asphalt.

 We are convinced that, under proper conditions, dew-ponds can be successfully made, and that in certain waterless districts with a moist atmosphere, their adoption would prove of inestimable advantage. Districts exist in which the air is moist, but the ground is always hot and the moisture finds no chilled surface as a foothold upon which it can be deposited.

Consider, for instance, the case of the three islands known as the Desertas. Their nearest point is about twelve miles from Madeira, and they are close together : the largest measures about ten miles by two miles. As their name implies, they are desert islands on which we believe that rain never falls. They are streamless and springless and uninhabited, except by a few fishermen who live on the largest. These men live by water obtained by spreading out fleeces in the evening, and wringing the dew from them in the morning.

If on one of the heights in the Desertas a dew-pond of good size were constructed, a plentiful supply of water should be obtained from the moisture-laden atmosphere. Thus vegetation could be supported, and if quickly growing trees could be nursed into existence, rain would be induced to fall and a total change in the climatic conditions would result. We fail to see why the Desertas should not in time be­come as fertile as Madeira itself—one of the richest spots on the face of the globe.

We are aware that we have only been dealing with the fringe of a fascinating subject. If any of  our readers care to go into the matter more fully with us we will gladly place further information at their service. For ourselves, we are not prepared to meet the great cost which is involved in a thorough investigation.


From Neolithic Dew-Ponds and Cattleways by AJ & G Hubbard 2nd edition 1907

Home ] Up ]

Copyright © 2002-6  Martin Snow and contributors as noted. All rights reserved.
Last modified: February 25, 2006