One other passage from the main cavern was explored, with a curious cluster of vertical cavities near its extremity. The end of the passage was coated in every direction with tinted deposits, among which we noticed beautiful specimens of the branching stalactites that were called _anemolites_ by the explorers of the Blue John caverns, who thought they had acquired their abnormal shapes through the irregularity of evaporation caused by air currents. I climbed 30 or 40 feet up one of the openings in the roof, whilst Dr. Sheldon explored another. At the top we found no exits big enough to afford a man passage. A wider cavity in the middle of the roof looked more promising. A ladder was adjusted, but fell short; but my companion, with considerable risk of a dangerous fall, clambered up to the rocky slope and over the piles of jagged blocks that well-nigh filled it. This too failed to afford us a passage, and the daring climber had great difficulty in coming down, being forced to thread the rope and let himself down on it to the ladder. During the operation a flake of rock came hurtling down and hit the ladder, but luckily did nothing worse than smash a rung. These cavities in the roof were extremely interesting, and no doubt are connected together and have a common origin in some neighbouring fissure or waterway. FIVE CAVERNS AT CHEDDAR The Cheddar gorge, which is the deepest and narrowest defile, and on its south side presents the loftiest face of absolutely vertical rock in England, is not dissimilar, though far superior in height and grandeur, to the Winnats pass in Derbyshire. The huge chasm runs east-north-east across the dip of the Limestone beds, which are tilted up towards the saddle of Mendip; one of its sides, consequently, is formed mainly of shelving rock, and the other is almost continuously precipitous. If, as may be assumed with confidence, the original cause of the ravine was a stream or streams flowing through a chain of caverns, one would naturally expect to find openings on the abrupt side through which the underground waters were successively tapped, and followed the trend of the strata to a lower level. This view is confirmed by observation. Except at the jaws of the defile, where both sides are equally high and precipitous, there are no caves on the northern side, but on the south openings both large and small are frequent, some narrow and lofty--"slitters," they are called locally--the others low and wide, according as they originated in a vertical joint or a bedding plane. They occur at various levels, some on inaccessible shelves high up in the cliffs, others along the base. But the larger number of these openings have in the lapse of time become silted up with clay and débris, so that the entrance is either completely masked or it is impossible to penetrate far without toilsome work with pick and shovel.