§ 6. According to the structure thus ascertained, the body of the shoot may at any point be considered as formed by a central rod, represented by the shaded inner circle, _a_, Fig. 36, surrounded by as many rods of descending external wood as there are leaves above the point where the section is made. The first five leaves above send down the first dark rods; and the next above send down those between, which, being from younger leaves, are less liable to interstices; then the third group sending down the side, it will be seen at a glance how a spiral action is produced. It would lead us into too subtile detail, if I traced the forces of this spiral superimposition. I must be content to let the reader peruse this part of the subject for himself, if it amuses him, and lead to larger questions. § 7. Broadly and practically, we may consider the whole cluster of woody material in Fig. 36 as one circle of fibrous substance formed round a small central rod. The real appearance in most trees is approximately as in _b_, Fig. 36, the radiating structure becoming more distinct in proportion to the largeness and compactness of the wood.[4] [Illustration: FIG. 37.] Now the next question is, how this descending external coating of wood will behave itself when it comes to the forking of the shoots. To simplify the examination of this, let us suppose the original or growing shoot (whose section is the shaded inner circle in Fig. 36) to have been in the form of a letter Y, and no thicker than a stout iron wire, as in Fig. 37. Down the arms of this letter Y, we have two fibrous streams running in the direction of the arrows. If the depth or thickness of these streams be such as at _b_ and _c_, what will their thickness be when they unite at _e_? Evidently, the quantity of wood surrounding the vertical wire at _e_ must be twice as great as that surrounding the wires _b_ and _c_.