Glacière caverns may, for the sake of convenience, be classified into several classes, according to their position or to their form. The lines of transition between them, however, are so indefinite in nature, that it is often difficult to specify a cavern as belonging to any special type. The most important factor in classifying glacière caves is their position. Under this head there are two main divisions: first, pit caves; second, cliff caves.

Pit caves are those where a pit or pits open into the ground, and the ice is found at the bottom. Sometimes there is no roof, when the place may be called a gorge: this occurs at Ellenville, where the roof has fallen. Again, the pit itself is more or less roofed over and the ice is found mainly or wholly under the roof: this is the case at Haut d’Aviernoz, at the Friedrichsteinerhöhle, at Saint-Livres, and at Saint-Georges. Sometimes the pit takes the form of a descending tunnel, leading into a hall or chamber, in which the ice lies under a rock roof: this happens at Chaux-les-Passavant. In all these pit caves the body of the cave is below the entrance, and most of them are fairly well lighted by daylight throughout. Generally there is only one pit, but occasionally there are two connected underground, as is the case at La Genollière.

Cliff caves are those where the entrance is at the base or in the side of a cliff. Frequently the cave is in the shape of a hall or chamber, which begins directly at the entrance, and which may be large or small. This kind always has a down slope directly from the mouth. The Kolowratshöhle, Dóbsina and the Grand Cave de Montarquis may be mentioned as examples. In some cases there is a pit at the base of a cliff and there is a slope leading down to the cave, somewhat in the form of a tunnel: this is the case at Manchester and practically also at Roth. Again there is a more or less long gallery between the entrance and the glacière, which is always below the level of the entrance. The Schafloch, Démenyfálva and Decorah may be cited as examples. As a rule the gallery slopes down from the entrance, but sometimes the floor rises and then sinks to the glacière. The top of the entrance, however, is always higher than the highest point of the floor, as otherwise the cold air could not get in. This is the case at the Frauenmauerhöhle, and, apparently, also at the Posselthöhle. In one case, at Amarnath in Kashmere, the floor is said to rise to the roof at the back; but as the entrance is nearly as big as the floor area, the ice formations must also be below the level of the top of the entrance.

The dimensions of glacière caves vary greatly. Some are large, others are small. Saint Georges, a roofed pit cave, is some twenty-five meters by twelve meters, with a depth of about twelve meters. Chaux-les-Passavant, a cave at the end of a pit tunnel, has a diameter of some twenty-seven meters. The measures of Dóbsina, a cave at the bottom of a cliff, are given as follows: Height of roof above ice floor, 10 to 11 meters; length 120 meters; breadth, 35 to 60 meters, and surface about 4644 meters. The Frauenmauerhöhle is a gallery about one hundred meters long before the ice floor is reached, and this is some fifty meters more in length by about seven meters in width. The glacière cave near Frain, on the contrary, is so small that one can only crawl in some two or three meters. In fact, glacière caves vary in size between great halls and little tunnels where one cannot stand up straight.

The entrances of glacière caves also vary greatly in their dimensions. For instance, the Friedrichsteinerhöhle is on one side of a huge pit and is as large and deep as the pit. Saint Georges, on the contrary, has, near one end of the roof, a couple of holes, some three meters in diameter. The entrance to the Schafloch is four meters wide by four meters seventy centimeters high, while the entrance to Roth is not over one meter each way.

A classification of subterranean ice formations, and one which applies to all the different forms, is into permanent and periodic glacières. When in any underground spot, ice remains throughout the year, the place may be called a permanent glacière; when on the contrary the ice melts away for part of the year, the place may be called a periodic glacière. This classification, which several observers have used already, is convenient and valuable.

Movements of Air.—Another classification of glacières can be made in accordance with the movements of air underground. Glacières may be divided into those where there are no strong draughts in summer and those where there are draughts: or into “apparently static caves”; and “dynamic caves” or “windholes.” The first class includes those caves where there is one or more openings close together and those above the body of the cave. In such hollows the air in summer is nearly still, while in winter there are distinct rotary movements of the air as soon as the temperature outside is lower than that within. Almost all glacière caves belong to this class of caves without strong draughts in summer. Sometimes, however, ice is found in hollows where there are two or more openings, at different altitudes and at different ends of the hollow, and where there are draughts. Occasionally, also, there are fissures in the sides or rear of apparently static caves, which allow something like draughts at times, as is the case at the Grand Cave de Montarquis.

Professor Thury of Geneva coined the terms “static cave” and “dynamic cave” which have come largely into use since, and which practically correspond to the German terms eishöhle and windröhre. I do not think the term “static cave” accurate, and prefer the term “apparently static cave” or “cave without distinct draughts.” For although there are many caves where the air seems stagnant at times, and there are no distinct perceptible draughts, still that the air is really stagnant all summer appears to me doubtful, and it seems as if the movements of air were distinctly apparent only in certain caves and not in others. Air which is apparently stagnant is found in both pit and cliff caves mainly in the summer months, but even in these I have noticed several times in summer slight movements of air, especially near the entrance. I could not exactly feel the air moving, but by lighting a cigar the smoke could be seen borne outwards exceedingly slowly. At the entrance of the Kolowratshöhle I think there was a faint outward current when I was there. The day was hot and windless, and as the cold air met the hot outside air it formed a faint cloud or mist at the mouth of the cavern. At Saint-Georges, although the air seemed tranquil, I found that the smoke of my cigar ascended rapidly just below the hole in the roof, showing an ascending air current. In the double cave of Chapuis, I found one cavern filled by a little lake over which there was a draught.[25]