I arrived at Wilsenroth on the 26th of July, 1897, and soon found an old forester, who said he had lived in the neighborhood for over fifty years, to show me the way. The Dornburg is a low hill, perhaps a hundred meters high and a kilometer long. It is basaltic and covered with sparse woods. The forester said that on top were the remains of the foundations of an old castle, and that this was possibly the origin of the name Dornburg. We circled round the eastern base of the hill for some ten minutes, when we came to a little depression, filled with basalt debris, among which were several small holes, out of which came currents of cool air.

Ten minutes further in the woods, we arrived at the Dornburg Restauration and then almost immediately at the glacière. It is at the bottom of a talus of broken basaltic rocks and has been much affected by the agency of man. In it are two eislöcher or stollen, as the forester called them. These are little artificial pits or cellars, dug into the talus. They are side by side, opening about southeast, and each is about one and a half meters wide, three meters long, and two meters high. The sides are built up with wooden posts and overhead is a thick roof of logs strewn with dirt. The day was cool and at the mouth of each eisloch, a faint outward current of air was discernible at nine-thirty A. M. I could not find any currents coming into the eislöcher. Inside it was cold and damp, and evidently thawing. There was a good heap of ice in each eisloch; it was clear, and I could detect no trace of prisms.

By much questioning, I dug out something of the history of these stollen from the forester. Formerly the ice was found at this spot, among the boulders at the base of the slope. But the people gradually took many of these basaltic blocks away, to break up for road making, and then the ice diminished. About 1870, a brewery, since burnt, was built at the Dornburg and the brewer had these stollen built, a sort of semi-natural, semi-artificial ice house. Every winter, the present owner of the stollen throws a quantity of snow into them, and this helps materially in forming the mass of ice.

Just below the restaurant there is a spring, which was said to be extremely cold, but there was nothing icy nor apparently unusual about it.

Under the restaurant itself is an interesting cellar. It was closed by wooden doors. First there was a passage way which turned steadily to the right, and which we descended by some ten steps. This was about two meters wide and was full of beer bottles and vegetables. On the left of the passage was a large double chamber where meat is kept. At eleven-thirty A. M. a faint draught blew down the passage and into the hall, the outside door being then open. The double hall was perhaps six meters each way, and I could detect no air currents coming into it at any place, except from the passage way. Both passage and halls were, as far as I could see, entirely built over with masonry. There was no ice and the temperature was some 7° or 8° above freezing point.

The daughter of the proprietor of the restaurant said that ice began to form in the cellar in February and that it lasted generally until October; but that this year it was destroyed early because the masonry was repaired, although it was still possible to skate in the cellar as late as March. In the beginning of winter the cellar was warm, and as she expressed it, der Keller schwitzt dann, which I suppose means that the walls are damp. She also said that it was a naturlicher Keller, and I am inclined to think that it was a natural glacière, converted into a cellar.

This visit to the Dornburg gave me many new ideas about classifying glacières, especially in relation to the movements of air. I was long puzzled by the German terms, Eishöhlen and Windröhren; and it suddenly struck me, at the Dornburg, that this terminology is incorrect, when used as a classification of glacières. The presence or absence of strong, apparent draughts, cannot be considered as a test as to whether a place is or is not a glacière; the presence of ice, for at least part of the year, alone makes a glacière, and this it does whether there are or are not draughts. It seems to me more than ever clear, however, that it all depends on the movements of air, as to whether ice forms in a cave. If the movements of air take the cold air of winter into a cave, then and then only—provided there is also a water supply—do we have ice. I am now inclined to think that caves, as far as their temperatures are concerned, should be classified into caves containing ice, cold caves, ordinary normal caves, and hot caves, without reference to the movements of air.

THE GLACIÈRE DE SAINT-GEORGES.

From Rolle, on the north shore of the Lake of Geneva; an excellent carriage road leads in two hours and a half to Saint-Georges in the Jura. At first the way goes steeply uphill and passes through many vineyards, and afterwards it crosses level fields to Gimel, then rises through woods to Saint-Georges. On arriving there on the afternoon of August 3d, 1897, I found the street filled with evergreens, and long benches and tables; the débris of a fête de tir, which had lasted for two days, with dancing and banquets and, I suspect, much vin du pays.

When I got down stairs at six o’clock next morning, all the people of the inn were sound asleep recovering from the effects of the fête, and instead of their calling me, I had to call them. Finally I succeeded in getting breakfast and then started in company with a first rate fellow, named Aymon Émery.