in structure, my analysis of which proved it to be andesitc. Some of the anorthosites described by Dr. T. Sterry Hunt in the Geology of Canada, 1 8G3, were proven by his analysis to be composed of pure labradorite, and some sections of the same which he submitted to me for examination were found to be composed of a multitude of small grains, none of which were twinned. Some of the tine crystals of oligoclase from Bodenmais are simple crystals so far as the ordinary mode of twinning is

more »

... If feldspar habitually showed their cleavages in their sections? the optical method might still be followed with some certainty, but as they do not, when the grains arc too small to allow cleavage fragments to be obtained for optical examination, the method followed by me* in the examination of the feldspathic constituent of the Triassic diabase is the most reliable. ' , In consideration both of the complexity of the feldspathic element in most rocks, and of the possibility of the simplicity of structure in triclinic feldspars, the very carefully developed methods founded upon the relation of twinning planes and elasticity planes in chance sections are liable to lead to wrong results. In volume III of the Proceedings of the United States National Museum, pp. 157-1G2, five species of Cretaceous fossils (together with some Tertiary species) were described, but not then illustrated. Illustrations of those Cretaceous species are now given on the accompanying plate of this volume, together with those of two other Cretaceous forms which are for the first time described in this article. The Arkansan species were collected by Mr. E. O. Ulrich in the vicinity of Little Rock, and by him presented to the Museum, together with a parcel of other fossils, mainly mollusca, which he found associated with them. The greater part of these Arkansan specimens are in the condition of mere casts of the interior of the shells, and therefore the determination of their specific and generic relations is not entirely satisfactory in all cases.