small that we could not advance another yard. My companion was some distance behind with his candle out, and I would not make a move until he had got it relighted, the consequences of both candles going out at once being unpleasant and possibly dangerous. For a long way we could not turn round, and had to crawl feet foremost. Just after repassing the junction my companion shouted that we were going wrong. He did not recognise the passage. I remained at the junction whilst he went farther and ascertained that it was the right channel after all. Then I examined the branch. It ascended 20 feet and then divided, the left branch, which was earthy, plainly striking up to the surface, the right branch going back towards the swallet. Undoubtedly there must be quite a labyrinth of dry water channels to correspond with the numerous series of openings in the cliff, but the one we explored seemed to be the largest and most practicable. Very tired and hot, not to mention the dirt, we made our way back to the exit, glad to feel that our day's work was done. The one thing that had impressed us most during our explorations was that all these swallets and water channels are cut through solid rock. Only when the rocks are shattered or disintegrated, as in the cases alluded to in Somerset, would there be any possibility of enlarging a swallet artificially. And though we had penetrated to a distance of 400 feet at Manifold we had not found the passages growing more roomy nor enlarged by the accession of tributaries. So far, the prospect of opening up the large fissures and chambers that must surely exist deeper in the rock seems unfavourable, unless the main channel of Giant's Hole can be unblocked. E. A. B. EXPLORING NEW CAVES IN DERBYSHIRE