It is clear that Croce was infinitely more interested in what Marx had actually accomplished, than in what he had intended to do. In Marx's own mind, the analysis of the conditions of capitalist society led inevitably to the conclusion that a passage from capitalism to socialism was predetermined by the structure of capitalist society itself. It is well known that the prevision assumed what claimed to be a strictly scientific character in the formulation of the law of the fall in the rate of profits: the gradual decrease of surplus-values accompanying the increase in technical improvements, and automatically re-establishing the equivalence of labour and value. Croce offered a very convincing criticism of this law on Marx's own grounds, by showing that it rested on a confusion between technical and economic facts, thus affording a remarkable example of Marx's uncertainty of his own method. It was clear that in this particular case Marx had been carried away by his desire to reduce the metaphysical implications of his economic sociology to the status of an historical law. And Croce's criticism was evidently intended both to deprive Marx's historical determinism of one of its most powerful instruments, and to confirm his own view of the method which gave validity and importance to such economic speculations as Marx's were. Marxian economics stood thus interpreted as comparative sociological economics, and by the definition Croce also defined the scope of sociological science, and the nature of the logical processes which it could legitimately employ. It was a considerable advance in the study of scientific concepts, as distinct from purely speculative concepts. But he still believed at the time, misled by the economists' discussions on the nature of the economic principle, that pure economics and the philosophy of economics practically coincided; and as he had maintained the legitimacy of Marx's method against the criticisms of the pure, or scientific, economists, he defended pure economics against Marx and his school, as the general science of the economic datum. But he was soon to understand that his own point of view and that of the pure economists were widely divergent, and that the methods of pure economics are in fact scientific and not philosophical. In later years, when he came to regard the science of economics as an empirical and mathematical science, Marxian economics appeared to him merely as a special branch of economic casuistry, employing methods fundamentally identical with those of pure economics: a relationship which could be illustrated by a comparison with the parallel of non-Euclidean and classical geometry. Although he did not reach this final position at the time, there is no doubt that his experience of the actual scientific processes of economics freed him from his allegiance to Herbartian logic, in which science and philosophy were still formally undifferentiated, and led him gradually