It seems to be generally accepted that four methods of expression are to be found in prose: _narration_, _description_, _exposition_ and _argumentation_. Narration deals with things in action, description with the appearance of things, exposition explains the relations ideas bear to one another, and argumentation not only does this, but tries at the same time to convince. Theoretically, this distinction is very easy to make, for action is the life of narration, appearance the theme of description, explanation and exposition are synonymous, and no one argues but with the hope of convincing. What can man do more than to tell what has been done, tell how a thing looks, show how one thing follows from another or is related to it, and endeavor to bring another person to the same state of mind? The accuracy and completeness of the classification is most evident until one attempts to apply it practically to existing literature, and then he finds that no literary masterpiece belongs entirely to any one of the classes, but that these mingle and unite, one or the other usually predominating. This ruling element, the one which is proportionately greater, will govern the classification of a selection. In any story, narration and description meet at every turn, and not infrequently exposition is found freely intermingled; while novels have been written with the avowed sole purpose of changing the beliefs of a people. _Uncle Tom's Cabin_ is a story of intense dramatic activity, and abounds in vivid descriptions of places and persons. It is generally dealing with incidents relating to the characters of the story, yet it really makes an exposition of the evils of slavery, and certainly was no small factor in stirring the American people into vigorous action against the slave dealers. Yet no one would classify the book otherwise than among the narratives. Although into Burke's _Conciliation_ other elements enter, yet everyone will admit it to be argumentative in the highest degree. So while it is well to classify the selections read, yet fine theoretical distinctions should be abandoned. It is not so necessary to classify and name as it is to compare and distinguish.