(b). The second method is the textual, and is especially though not exclusively adapted to sermons. In it a verse from the Bible, a motto, a sentence used by an opponent, or some definite form of very significant words, affords a basis for each part of the discourse. The order of the discourse may, however, be different from that of the words in the text, any change being allowable which secures more of the advantages of the narrative or logical methods. When the text is itself well known, a plan based upon it has an obvious advantage in assisting the memory both of speaker and hearer, by suggesting each part of the discourse at the proper time. When any lecture or oration has a formal motto which sums up and fairly expresses the subject discussed, the textual plan will be as well adapted to it as to a sermon. (c). The logical or mathematical method is the third and probably the most symmetrical form the plan may assume. A topic is taken, and after the introduction, which may be the mere statement of the subject, or of the relations of the speaker or of the audience to it, that subject is unfolded with all the precision of a proposition in geometry. Each thought is preliminary to that which follows, and the whole ends in the demonstration of some great truth and the deduction of its legitimate corollaries. This method is the best possible in those cases adapted to it--particularly those in which some abstruse subject is to be unfolded and proved. (d). The last method we will describe proceeds by divisions and subdivisions. It is the military method, for in it the discourse is organized like an army, into corps, brigades, and regiments; or it is like a tree, which divides into two or three principal branches, and these again subdivide until the finest twigs are reached. All the detached items that have been selected are brought into related groups each governed by a central thought, and these again are held in strict subordination to the supreme idea. A subject will many times arrange itself almost spontaneously into several different parts, which thus form the proper divisions, and these again may be easily analyzed into their proper subdivisions. Even when this is not the case, we will see, as we examine the jottings we have made while gathering our materials, that a few of the ideas stand out in special prominence, and with a little close study of relations and affinities all the others may be made to group themselves around these. The individual ideas we put down on the first study of the subject usually form the subdivisions, and some generalization of them. It is not well to make the branches of a subject too numerous or they will introduce confusion and fail to be remembered. From two to four divisions with two or three subdivisions under each, are in a majority of cases better than a large number. The tendency to multiply them to a great extent, and then to name them in