38 SCIENCE AND FRUIT GROWING 



but in their place there has developed a new root-system from the 

 buried portion of the stem, and these new roots, having been 

 unchecked in their growth since first coming into existence, have 

 developed exceedingly, so much so that their weight proved to be 

 over three times that of the roots of the stocks planted at a depth 

 of six inches ; and the branch growth, consequent on this large 

 root-formation, showed almost as great an increase. A series 

 of experiments with stocks planted 12 inches below the surface 

 gave results intermediate between those at 6 and 24 inches. 

 These results were obtained at Harpenden, where the soil is 

 favourable to the growth of young trees, but similar, though less 

 conspicuous results were obtained at Ridgmont. On the other 

 hand, a series made in the light sandy soil at Millbrook gave 

 ambiguous results, for though the deeply planted stocks were 

 superior to the others as regards the weight of roots and branches 

 formed, the total weights of the trees on lifting showed a de- 

 ficiency. It is evident that the optimum depth below the 

 surface for the development of roots is mainly conditioned by 

 aeration, and in a sandy soil, where aeration extends to a great 

 depth, differences between deep and shallow planting would 

 be less conspicuous. This was shown by some earlier experi- 

 ments on the subject (V, 69), in which some of the deeply planted 

 stocks were planted in ground with the layers of soil inverted, 

 that is, with the upper nutrient layers placed round the deeply 

 buried roots, and the clay subsoil round the stems : but the 

 original roots, in spite of their rich environment, did not develop, 

 whilst the stems, in spite of their unfavourable surroundings, 

 sent out strong new roots, though not so strong as in the case 

 where the good soil was kept at the surface. 



The paradise stock, which was used in these experiments, 

 develops new roots very easily, but that is not so with the crab 

 stock, and similar experiments made with this led only to a great 

 decrease in growth when the depth of planting was increased, 

 the increments in the tree weights after two years being in the 

 proportion of 100 : 61 : 33, when the depths were 6, 12 and 24 

 inches, respectively. Even in the case of paradise stocks it is 

 evident from an inspection of Fig. 6 that the extent of the new 

 root- development varies very much with the individual plant. 

 Subsequent experiments with quince stocks showed that they 

 behaved in the same way as crab stocks, and not like paradise 

 stocks, as had been expected. 



One feature of these results requires special mention, namely, 

 that, in all cases where the buried paradise stocks eventually 



