J 53 The Cordon System of Training Fruit Trees. 



several times noticed it in good condition j but generally the pear 

 pushes a little too vigorously to be trained in that way, while the 

 pendulous habit of the fruit renders it more liable to be soiled. I 

 once saw Uvedale's St. Germain pear grown in this way, the great 

 fruit sitting on the ground, and quite encrusted with earthy splash- 

 ings. However, some of the best of the free-bearing kinds I have 

 seen used to better purpose as edgings, and know no reason why first- 

 rate varieties of compact fertile habit cannot be grown well in this 

 way, especially on the fruit borders, where they might be protected 

 so efficiently. Planted on a border devoted to peach trees, the same 

 protection might easily be extended to them ; and thus we might, 

 by paying a little more attention to securing perfect protection, get 

 splendid crops of both fruits. 



The Peach as a Cordon. — ^With the peach as an oblique 

 cordon, a better result is attained than with the pear, the wall 

 being covered very rapidly, and the neat laying in of a great number 

 of shoots on each side of the simple stem does away with the 

 crowded and unnatural appearance which a plantation of cordon 

 pears assumes when old and the stems are thickened. I have seen 

 some really good results produced in this way, both in glass-houses 

 and against walls. 



But instead of the wood being pinched in, as people might suppose 

 in England from reading of the method of one Mr. Grin, it is nailed 

 or tied in at each side of the branch, aye, more so than if that branch 

 were part and parcel of one of the older and larger forms of tree. I 

 once saw an excellent result aflbrded by this system against the high 

 back wall of a vinery in the establishment of M. Rose Charmeaux, 

 at Thomery. By its means he perfectly covered his wall in a short 

 time, and gathered a great variety of fruit from a small space. Out 

 of doors I have seen it afford beautiful results, and that not un- 

 frequently. It is well calculated for high walls, and it may be 

 adopted for low ones by training the trees at a more acute angle 

 with the earth. Fig. 33 shows how the ends of a wall thus covered 

 with cordons are to be furnished. Considering the time usually 



