2 66 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



a better constitution than the French, hence it may have a very 

 special use. 



It seems certain that either Type IV (the Holland stock) (fig. 54) or 

 Type V (Improved Doucin) (fig. 56) would serve well as a dwarfing 

 stock for garden purposes, especially for cordons. 



A very complete series of Paradise would be made up by the 

 selection of : 



Type 1. Broad -leaf — for permanent bush- trees and weak-growing 

 sorts. 



Type II. Doucin or English Paradise — for bush fillers and general 



commercial dwarfing trees. 

 Type V. Ameliore or Improved Doucin — for cordons and garden 



trees. 



Type IX. Jaune de Metz or Yellow Paradise — for extra precocity, 

 probably useful on which to work new seedlings, pot trees, &c, 

 to induce rapid maturity (fig. 55). 



• 



Such are, at any rate, the most marked and constant indications 

 on our soil with regard to the definitely dwarfing types of stock. 



Finally, I must say a word about the free-stock problem. We 

 have seen that there is no distinct line betwixt free and Paradise. 

 Each contains dwarfing and free-growing stocks, both of which 

 are capable of vegetative propagation. By this latter means only 

 can any particular type be kept true. If, then, we are to eliminate the 

 patent inequalities in our plots of standards in the nursery and in the 

 orchard, we must by some method standardize our free stock, instead 

 of either burning a high percentage of culls or turning them into 

 inferior cordons or bush-trees. There are three possible lines along 

 which this standardization may proceed. In the present stock crisis 

 it seems desirable that all three methods should be used. 



1 . Whilst seedling stocks are to be used they could be graded at an 

 early stage, i.e. at the time of planting out, along the lines indicated 

 in the Bristol Report on Free Stocks. It is probable that both their 

 earliest groups and their final ones will produce dwarfing rather than 

 free-growing stocks, and should therefore be discarded or used solely 

 for dwarfing purposes. Our experience so far has demonstrated that 

 a fairly high percentage of free stocks will fall in groups C, F, and G, 

 apparently all more or less free-growing. It is impossible to grade these 

 stocks from early vegetative characters alone with any degree of 

 accuracy. The roots must be taken into account at the initial grading. 



I am inclined to think that, in any case, this can be only a temporary 

 and passing measure. All the same, it is worth initiating at once. 

 And it could be followed up by a de-roguing of our free-stock beds 

 of obvious weaklings before the budding and grafting season. 



2. It is possible amongst our stronger types of so-called Paradise 

 to pick out those, such as X (fig. 59), XIII (fig. 61), and XVI, with 

 free-rooting and growth habit and deep anchorage, which appear 

 identical with free stocks. Any of these types can be raised quickly 



