Our study has been confined almost entirely to the feldspars of routine rocks and has been directed away from the feldspars of pegmatites. Actual separations were made magnetically and by heavy liquids, largely by Gates and Clabaugh (Chapter 1), and the fractions were chemically analyzed (Chapter 2). To the extent that our analyzed feldspars are representative of all such rocks, we conclude that the potassic content of plagioclase is much lower than we have been led to believe; also, that perthitic feldspars are probably composed of members which are quite normal in composition as defined by the analyses of the feldspars given. We believe that soda-rich “orthoclase” will prove to be as difficult to find as is potash-rich plagioclase. The optical study of our samples and the correlation with the analyses were done by Crump and Ketner (Chapter 3). They have learned that the refractive indices of plagioclase give the most reliable information on the chemical composition. The potash feldspars have not yet been studied in the same way. The Fedorow migration curves are less reliable in our present state of understanding. We hope that the irregularities they display may ultimately prove to be their greatest asset in that, when understood, they will yield even more detailed information. That will be our next endeavor. Extinction-angle curves, properly selected and used, are apparently more reliable than are the Fedorow migration curves but are less reliable than the refractive indices. We have made no progress toward a recognition of the potash content of plagioclase from optical data. We believe the optical irregularities that we have detected are inherent within the crystals and cannot be explained yet by their temperatures of formation since the range of irregularities may be found entirely within one sample. Feldspar zoning reflects some aspects of the environment under which the zoned crystal formed. Zoning and polysynthetic twinning in plagioclase tend to be mutually exclusive. Polysynthetic twinning, which forms after the crystal is nearly or fully grown, must follow zoning. Since zoning disappears as twinning develops we conclude (Chapter 4) either that the process of twinning eliminates zones or that the environment which favors twinning is unfavorable to the continuance of zones. Very mildly twinned crystals are apt to be zoned. Heavily twinned crystals are occasionally zoned but commonly are not. Some crystals show zoning in one portion and twinning in another. The full story is not known since the observation includes many exceptions. Some twin composition faces follow zones which parallel crystal faces. That is, some twin laws are predetermined by the zoning. There is some support for the thought that the zoning in general determines the polysynthetic twin law to follow. Potash feldspar may contain a large amount of plagioclase feldspar in solid solution at the temperature of formation of the feldspar. Most of this plagioclase is unmixed on cooling, the amount unmixed to be determined when the potash feldspars are similarly studied. Gates has restudied perthites (Chapter 5) and has learned that the plagioclase feldspar unmixed may remain in its host in fine or coarse perthitic stringers or may move about to replace other orthoclase feldspar at points of. concentration. The movement of this mobile sodic feldspar appears to be controlled by pressure differentials brought about through structural disturbance. Under favorable conditions the sodic solutions may migrate considerable distances to accumulate in still greater concentrations to create peralkaline rocks (Chapter 6). At the relatively lower temperatures of deuteric solutions plagioclase yields its potash content to create sericitization effects and to render alkaline the solutions associated with it. These alkaline solutions, interstitially distributed through the rock, may dissolve iron and magnesium selectively from the mafic minerals. The resulting solutions may be traced to fractures where they produce lamprophyres (Chapter 7) or to other centers of accumulation to produce mafic segregations. Parenthetically, we report that radioactive material follows a parallel route, some of it being deposited as pitchblende in close association with lamprophyres. Bradley and Lyons describe (Chapter 8) the reduction in grain size of phenocrysts to normal granitic textural size by the recrystallization of plagioclase. Hence at least some granites appear to have originated through the route of greenstone metamorphism, first to an augen gneiss, then by transition to a coarse feldspathic porphyry, and finally to a normal granitic texture and composition by the recrystallization of the phenocrysts to a feldspathic aggregate only slightly disturbed. Finally, criteria are offered (Chapter 9) designed to help recognize those granitic rocks in which the critical role of fluids during crystallization have left their trace. The so-called Wisconsin batholith has been studied from this viewpoint and is interpreted tentatively on this basis. Some observations are also made on the Canadian shield. We conclude that some granites have formed with little liquid influence on their crystallization, others with a prominent liquid factor. The concentration of liquid has been very irregular and not constant at any one place. Overall, the vast majority of our granites would seem to be shallowlv exposed; our observations are largely confined to the roof regions.