PLANTING 35 



before the following spring, it was found that the production of 

 new rootlets had been decidedly greater in the case of the trees 

 with the mangled roots (IX, 52) ; whilst, on the other hand, in 

 some later experiments (XV, 25), where the question was in- 

 vestigated more systematically by planting some 60 apples, 

 pears and plums, on which the same number of roots of equal 

 size (J to J an inch in diameter) had been selected, and had been 

 broken in one case and carefully cut in the other, it was found 

 that the new rootlets formed from the old roots which had been 

 broken showed a deficiency of about 15 per cent, as compared 

 with those formed in the case where the roots had been cut. 

 But in another similar experiment with 50 apples trees of five 

 different varieties, the results showed practically no difference 

 between trees with trimmed or mauled roots. With currants and 

 gooseberries, however, the results were very definite, for the new 

 root-formation from the broken roots was two -and-a- half times 

 that from the cut roots, a superiority which was very surprising 

 and difficult of explanation (cf. p. 54). 



Thus, with one series giving results in favour of trimming, two 

 giving results against it, and a fourth giving no results at all, the 

 only possible conclusion is that trimming is altogether unim- 

 portant, and that the omission of it may sometimes have one 

 effect, sometimes another. 



The effect of the removal of the fibrous roots was investigated 

 in a separate experiment (IX, 46), plum trees being planted after 

 the removal of all rootlets up to I mm., 2 mm., and 4 mm. in 

 diameter in different cases. After two years' growth, the trees 

 when lifted showed by the increase in their weight that the 

 removal of the smallest fibres ( I mm.) had been beneficial to 

 the extent of 17 per cent. ; that the removal of the fibres up to 

 2 mm. had produced no effect ; whilst the removal of fibres up 

 to 4 mm. had produced a bad effect, amounting to a loss of 

 9 per cent, in growth thus affording another instance of detri- 

 ment to the tree not following injury till the latter exceeds a 

 certain limit. But the effect of the removal of the fibrous roots 

 must vary, as might naturally be expected, with the readiness 

 with which the tree will send out new rootlets, and hence with its 

 character and age : accordingly, in the case of some beech trees 

 of about ten years of age, it was found that the removal of fibres, 

 even up to only I mm. in diameter, had a markedly deleterious 

 effect. 



At the same time some pear trees were planted with their 

 roots all tied together in a ball (Fig. 7, D, p. 41), in order to 



