TRAINING. 197 



aronnd the stems of rider trees ; and the stellate fan is a capital form for 

 at once utilising part of this waste, while it is also beantifnl in itself. 

 Figs. 18 and 19 show stellate riders half and fully formed, either of which 

 may be employed according to circumstances or the space at command. 



Seymour's system of fan training (Figs. 20, 21, and 22) is that most 

 generally adopted. As a rule, two general features distinguish it from 

 the open or common English fan. One is that, generally in Seymour's 

 method, a centre leader is retained in the tree for the furnishing of sub- 

 leaders at regular intervals, and the other that all the fruit-bearing wood 

 is laid in on the upper side of the branches. No doubt the retention of 

 a central leader does much to insure a constant supply of branches. By 

 cutting this centre, which is generally the strongest shoot on the tree, 

 pretty severely back, it cannot fail to produce two, four, six, or even 

 more sub-leaders a season as needed, and at the points required. This 

 last is a matter of much moment, and it cannot always be relied on in 

 the open fan mode of training. 



Nevertheless, the central leader can hardly be designated as essential 

 to the adoption to what is, peAaps, the chief merit of Seymour's system. 

 That is, undoubtedly, the placing of all the bearing wood on the upper 

 side of the sub-leaders and other branches. The chief object of this 

 arrangement is to place all the bearing shoots, not only on an equilibrium 

 of advantage, but the whole of them in the best positions. That, it 

 has been assumed, and, perhaps, correctly, is the upper part of 

 the branch. The sap is supposed to flow to each with equal force and 

 volume when so placed. The trees, too, manifest more system, and dis- 

 play greater order trained in this way. IJnless, however, considerable 

 care is taken to furnish the upper sides of the leaders with rather more 

 bearing wood than in the ordinary mode of fan training, it is obvious 

 that the trees wiU really be less fully furnished with wood. This is 

 apparent, and is meant to be made so in the woodcuts. It is often, how- 

 ever, a decided advantage to have less wood, and overcrowding is one of 

 the most patent causes of the decline and fall of peach and other trees. 

 Another objection that has often been urged against Seymour's method is 

 the divergence of the side shoots at right angles to the central stem. 

 The vertical stem, which, of course, ultinlately grows into a trunk, is 

 very prone to appropriate the lion's share of the supplies, and the side^ 

 leaders, especially the lower ones, to be starved into weakness or wholly 

 perish from this diversion of sap to where it is seldom greatly wanted. 

 Fig. 23 furnishes an illustration of a simple means of neutralising the force 

 of this objection very much, if not of removing it altogether, alike 

 in the case of Seymour's and the common open fan trainiag. 

 The lower and other shoots are not only elevated, and kept so 



