These features, together with those predicted by Darwin, have been recognised by von Richthofen and Mojsisovics in the Tyrol dolomites, and have afforded Austrian geologists good evidence that large parts of these limestones originated as coral-reefs. Faulting, however, has undoubtedly taken place in this region, producing here and there a subsidence of the limestone blocks among the surrounding more normal sediments. Rothpletz, Ogilvie Gordon[10], and other critics of von Richthofen's view have seen in this faulting the cause of the abrupt change from a facies of massive dolomite to one of normal sedimentation on the same horizontal level. They have also urged that shell-banks may accumulate locally so as to simulate reefs by their contrast with their surroundings, while the change to dolomite has obliterated their original features (see p. 30). It cannot be denied, however, that coral-reefs and their associated detrital deposits must exercise a very important influence in the formation of solid limestone. Even small knots and local groups of compound corals are seen in ordinary limestones to serve as a mesh in which other organic remains have become entrapped. The ease with which the aragonite of their skeletons becomes silicified causes them often to stand out on weathered surfaces with all the delicacy of structure displayed upon a modern reef. Where limestones and shales are associated together, a "knoll structure" may be found, the limestone occurring in masses of a somewhat hemispherical form, with the shales fitted against and round them. In some cases this may be due to the local distribution of patches of growing coral on the old sea-floor; but in other cases the structure has arisen from compression and brecciation of the strata, the original beds of limestone becoming broken up and the more yielding beds flowing round them. This structure is well seen on a small scale in many "crush-conglomerates," where the limestone appears as knots and eyes, resembling pebbles. Yet near at hand the true bedding may be traced, bands of limestone alternating with shale, and a few cross-joints indicating the possibility of a separation of the limestone into blocks. These blocks become rounded in the general rock-flow; but Gardiner and Reynolds[11] suggest solution by infiltering water as an explanation of certain remarkable examples studied by them. ALTERED FORMS OF MASSIVE LIMESTONE