He provisionally sets aside others with a different connexion, though without for that reason suppressing them, intending rather to recall them when another problem makes their presence necessary. It is for this reason that even in those abridgments that seem to be the most complete and 'objective' (that is to say, 'material') a certain selection does appear, due to the theoretical interest of the writer, who never altogether ceases to be a historiographer-philosopher. The procedure is in fact just that of the history of art and poetry, where what is really historical treatment, living and complete, is the thought or criticism of individual poetical personalities, and the rest a table of criticisms, an abridgment due to contiguity of time or place, affinity of matter or similarity of temperament, or to degrees of artistic excellence. Nor must we say that every philosophic problem is linked to all the others and is always a problem of the whole of philosophy, thus differing from the cases of poetry and art, for there is no diversity here either, and the whole of history and the entire universe are immanent in every single work of art. Now that we have likewise reduced philosophies of history to the rank of particular histories, it is scarcely necessary to demonstrate that the demand being made in several quarters for a 'universal' or 'general' history of science is without foundation. For such a history would be impossible to write, even if we were able to identify or compare the history of science with chat of philosophy. But it is doubly impossible both because there are comprised under the name of 'science' such diverse forms as sciences of observation and mathematical sciences, and also because in each of these classes themselves the several disciplines remain separate, owing to the irreducible variety of data and postulates from which they spring. If, as we have pointed out, every particular philosophical problem links and places itself in harmony with all other philosophical problems, every scientific problem tends, on the contrary, to shut itself up in itself, and there is no more destructive tendency in science than that of 'explaining' all the facts by means of a 'single principle,' substituting, that is to say, an unfruitful metaphysic for fruitful science, allowing an empty word to act as a magic wand, and by 'explaining everything' to 'explain' nothing at all. The unity admitted by the history of the sciences is not that which connects one theory with another and one science with another in an imaginary general history of science, but that which connects each science and each theory with the intellectual and social complex of the moment in which it appeared. But even here too we must utter the warning that in thus explaining their true nature we do not wish to contest the right to existence of tables and encyclopædias of the history of science, far less to throw discredit upon the present direction of studies, by