[(3) Divisions of philosophy.] Nor is it otherwise than in harmony herewith that logic did not receive a fuller or more perfect treatment in the further development of their system. Even the study of nature, going far more fully into particulars than logic, was guided entirely by practical considerations, all scientific interest in nature being ignored. Following the usual method, however, the Epicureans divided philosophy into three parts [1119]—logic, natural science, and moral science. Limiting the first of these parts to one branch of logic, the part which deals with the characteristics of truth, and which they therefore called neither logic, nor dialectic, but Canonic, they really reduced this part to a mere introductory appendage to the two other parts, [1120] and studied Canonic as a part of natural science. [1121] Natural science moreover was so entirely subordinated to moral science, that we might almost feel tempted to follow some modern writers [1122] in their view of the Epicurean system, by giving to moral science precedence of the two other parts, or at least of natural science. [1123] The School, however, followed the usual order, and not without reason; [1124] for although the whole tendency of the Epicurean Canonic and natural science can only, like the Stoic, be explained by a reference to moral science, yet moral science with them presupposes the test-science of truth and natural science. We shall, therefore, do well to treat of Canonic in the first place, and subsequently to prove how this branch of study depends on Ethics. [B. Canonic or the test-science of truth.] [(1) Sensation and perception.]