Canonic or the test-science of truth, as has been observed, is occupied with investigating the standard of truth, and with enquiring into the mode of acquiring knowledge. The whole of formal logic, the doctrine of the formation of conceptions and conclusions, is omitted by Epicurus. [1125] Even the theory of the acquisition of knowledge assumes with him a very simple form. If the Stoics were fain, notwithstanding their ideal ethics and their pantheistic speculations, ultimately to take their stand on materialism, could Epicurus avoid doing the same? In seeking a speculative basis for a view of life which refers everything to the feeling of pleasure or pain, he appealed far more unreservedly than they had done to sensation. Now, since the senses can alone inform us what is pleasant or unpleasant, and what is desirable or the contrary, our judgment as to truth or falsehood must ultimately depend on the senses. Viewed speculatively, sensation is the standard of truth; viewed practically, the feeling of pleasure or pain. [1126] If the senses may not be trusted, still less may knowledge derived from reason be trusted, since reason itself is primarily and entirely derived from the senses. There remains, therefore, no distinctive mark of truth, and no possibility of certain conviction. We are at the mercy of unlimited doubt. If, however, this doubt is contradictory of itself—for how can men declare they know, that they can know nothing?—it is also contradictory of human nature, since it would do away not only with all knowledge but with every possibility of action—in short, with all the conditions on which human life depends. [1127] To avoid doubt we must allow that sensation as such is always, and under all circumstances, to be trusted; nor ought the delusions of the senses to shake our belief; the causes of these deceptions do not lie in sensation as such, but in our judgment about sensation. What the senses supply is only that an object produces this or that effect upon us, and that this or that picture has impressed our soul. The facts thus supplied are always true, only it does not follow that the object exactly corresponds with the impression we receive of it, or that it produces on others the same impression that it produces on us. Many different pictures may emanate from one and the same object, and these pictures may be changed on their way to the ear or eye. Pictures, too, may strike our senses with which no real objects correspond. To confound the picture with the thing, the impression made with the object making the impression, is certainly an error, but this error must not be laid to the charge of the senses, but to that of opinion. [1128] Indeed, how is it possible, asks Epicurus, [1129] to refute the testimony of the senses? Can reason refute it? But reason is itself dependent on the senses, and cannot bear testimony against that on which its own claims to belief depend. Or can one sense convict another