The dolomitic facies of the Carboniferous limestone in our islands is an example of the second type of origin. The dolomite here frequently occurs in irregular veins and patches. The introduction of iron carbonate with the magnesium salt stains the dolomite brown on exposure to oxidation, and its limits are thus clearly seen in the general blue-grey mass. The dolomitisation has evidently proceeded from joint-surfaces inwards. It is often sufficiently thorough to obliterate all traces of fossils, and the shrinkage accompanying the chemical change has produced numerous cavities, in which calcite has subsequently crystallised. An expansion takes place when aragonite is altered into dolomite, unless more of the calcium carbonate is removed than is necessary to give place to the magnesium carbonate introduced. In the change from calcite, with a density of 2·72, to dolomite, with a density of 2·85, there is, on the other hand, a shrinkage of 4·56 per cent. Where the alteration, then, takes place while the aragonite organisms still remain as aragonite, and not as calcite, an expansion rather than a contraction should occur in the substance of a reef; but when an old limestone, in which all the calcium carbonate is present as calcite, becomes dolomitised, a considerable shrinkage will occur, and rifts and hollows may remain obvious. Very few dolomites, except those found in association with rock-salt and other products of the evaporation of lagoons, can now be attributed to direct chemical deposition from the sea. Daly[7] has argued that the first Palæozoic and the pre-Cambrian dolomites were formed by precipitation, since the calcium salts in those early days were completely removed from the sea-water. Ammonium carbonate, though effective in precipitating the calcium salts, does not act on those of magnesium until the calcium salts have been brought down. But, under the conditions postulated for the river-waters that reached the sea from the earliest continental lands, conditions involving the presence of only small quantities of salts of calcium, the decay of organisms on the sea-floor might lead to a deposition of all the magnesium salts, following on those of calcium, both coming down in the form of carbonates. The experimental work of Pfaff[16] should be considered in connexion with Daly's suggestions, since means are there indicated whereby basic magnesium carbonate, precipitated from sea-water, may associate itself with calcium carbonate to form dolomite; shallow-water conditions, with concentration by evaporation, are required. Daly compares analyses of river-waters now running over pre-Cambrian rocks with analyses of pre-Cambrian limestones, and the ratio of the carbonates of magnesium and calcium is shown to be the same in both series.