There are, however, two distinct notes in the color effects of glacière caves and these may be described as the partly subterranean, or as the wholly subterranean. In the former case the local tints stand out more clearly. For instance, at the Kolowratshöhle the ice is beautifully transparent and of a pale ochre-greenish hue: the limestone rocks are streaked with iron, and thus have a reddish hue, while, owing to the entrance admitting plenty of daylight, the effect is only semi-subterranean. Again, at Chaux-les-Passavant plenty of daylight is admitted: the rocks are a yellowish brown, and the ice is white and blue. At the Schafloch or the Frauenmauer, on the contrary, the effect is wholly subterranean: daylight is so completely absent that black is the predominating note, the ice itself looking gray. Dóbsina is an exception, as, thanks to the electric light, white is the conspicuous tone, even though rocks and shadows dull many places and corners into a sombre gray. More than once, on returning to daylight from the intense blackness of a cave, I have seen the rocks near the entrance appear a dark blue color, exactly simulating moonlight. This effect is common to both glacière caves and ordinary caverns. It is a striking but rare phenomenon, and depends apparently on the shape of the cave. This moonlight effect only seems to occur when a cave makes an elbow directly after the mouth and then goes straight for some distance. When the daylight is actually in sight, the moonlight impression vanishes. _Carbonic Acid Gas._--Carbonic acid gas, judging from the most recent explorations, is more of a rarity in rock caves with normal temperatures than is generally supposed. There appears to be only one case on record where this gas was observed in a cold cave. This was in the Creux-de-Souci,[33] which is rather a cold than a freezing cavern, but which on one occasion was found to contain snow, and whose temperature is always extremely low. From the present state of knowledge, therefore, it may be assumed that if carbonic acid gas does form in glacière caves, it does so only seldom. [33] See Part III.: page 207.