An average analysis is made of an altered intrusive greenstone occurring in the zone of the Kiruna greenstones with the special object of proving whether any local change of chemical composition is occasioned during the alteration processes by the uneven distribution of epidote, as was earlier supposed by the writer. The analysis confirms this, but it also shows that chemical inhomogenities must originally have occurred in the rock-mass. The relations found thus do not indicate great local changes of the chemical composition during the mineral alteration processes still less a change of the bulk composition of the rocks.

A comparison of a number of analyses of spilites and related An-richer rocks affords the following facts: among them there occurs a small group of An-rich members, which are designated here as diabases. Most of the analyses refer to rocks richer in albite, which are here called spilites in the proper sense. In the felspar diagram the projection points of the calculated felspar mixtures of spacing the rocks form a narrow zone close to the plagioclase line extending from 7 per cent An to 58 per cent An. As the dividing point between the diabases and the spilitic members a mixture with about 40 per cent An is accepted. When adding to the diagram the keratophyric rocks associated with the spilites, the whole distance from about 60 per cent of An to the immediate vicinity of the Ab-corner would be filled up.

Between the spilites and other rocks of equal femic composition and of about equal SiO2 content there exists in the felspar diagram an area in which very few rocks are represented. Petrographically the spilites are thus distinguished as a special rock-type. The boundary towards the ordinary basalts and diabases at the An-rich end of the series on account of the relations of the felspars alone is more dubious, as the individual basalts or diabases may be very poor in potash, but a difference seems to exist here also.

Other chemical qualities characteristic of the spilites are deficiency in alumina and high content of ferrous iron and of TiO2. The low percentage of Al2O3 is the main cause of the soda-rich composition of the felspar, as the CaO-content of the rocks is not low. The An-poor composition of the felspar in the rocks, furthermore, is accentuated by the entrance of An-molecules into the pyroxene during the crystallization of the magma, the generation of olivine being thereby prevented. This mineral, or pseudomorphs of it, is seldom recorded from the spilites. The high content of FeO is distinctive, not only for the spilites proper but also for the diabase members truly belonging to the series.

The importance of the high content of FeO for the course of crystallization in the An-poor members of the spilites is discussed. The consequence is found to be a change of the eutectic ratio of pyroxene and plagioclase contained in the concentration-diagram of plagioclase and pure MgO-pyroxene. The eutectic line for pyroxene and plagioclase during the crystallization of the magma will thus be reached earlier than shown by the diagram. In this way the textural relations of the rocks are explained without the hypothesis of autometamorphic changes and of an earlier An-rich plagioclase. In the appearance and mineral composition of the rock there is nothing compelling an assumption of this kind, unless the alteration of the minerals to compounds richer in water and carbon dioxide is referred to the period of consolidation. This would imply the fixing in the consolidated rock of these elements, but not a change of the other oxides. As to the time of the generation of the secondary minerals we do not know anything with certainty. The fact that the spilite group includes members which are poor in water and carbon dioxide, or are not decomposed at all, shows that a high content of these compounds is not necessarily inherent in the spilitic magma.