Food Adjuncts and Drugs.--In addition to the nutrients and the small quantity of indigestible fibre of which we have already written, food generally contains small quantities of substances which are difficult to classify, and whose action on the body is but imperfectly understood. Many of these possess pungent or strong odours and flavours. To them, various fruits, meats, etc., owe much of their characteristic differences of taste. When pure the proteids and starches are devoid of taste. Such oils and fats as are generally eaten have also but little flavour, providing they are free from rancidity and of good quality. The sugars differ from the other nutrients in possessing a more or less decided taste. The free vegetable acids also strongly affect the sense of taste, but they are only consumed in small quantities. A drug may be defined as a substance which modifies the functions of the body or of some organ without sensibly imparting nourishment. This action may be one of stimulation or of depression. A drug is taken for its medicinal action, a food adjunct for its modifying action on food. It is impossible to give a quite satisfactory definition, or to draw sharp distinctions. For example, tea, coffee, alcohol and tobacco are sometimes placed in one group, and sometimes in another, according to opinion of their action and the definition of the terms food adjuncts, drugs and poisons. The difference of grouping often depends upon intensity rather than of kind of action. If taken frequently and not in quantity sufficient to have a markedly medicinal action, such things are generally called food adjuncts or supplementary foods, although much may be said in favour of a different view. The volatile oils of mustard, caraway, cloves, etc., are used in medicine; also the alkaloids of coffee and cocoa. Even honey is used as a mild laxative for infants; that is, as a drug. The difference between a drug and a poison is one only of degree. Some of the most esteemed drugs have to be administered in very small quantities, or they cause death; e.g., strychnine and morphine. Classifications are necessary for methodical study, and for assisting the memory in grasping large numbers of things which can be grouped together. Classifications, however, are artificial, not due to natural lines of demarkation, but according to man's knowledge and convenience; hence a group is apt to approach and finally merge into another group, although on first consideration they appeared quite distinct. The disregard of this often leads to confusion and useless discussions.