54 SCIENCE AND FRUIT GROWING 



plants examined. Both series, however, agreed in showing that 

 the extreme ends of the old roots are not the chief seat from 

 which new rootlets arise, as is generally stated to be the case ; 

 and, even if we embrace under the term " extreme end " a 

 stretch of a quarter or half an inch from that end, these root- 

 ends still do not occupy a position of paramount importance as 

 seats of origin for new rootlets, especially when it is borne in 

 mind that the new roots originating from them are more feeble 

 than those originating from elsewhere (p. 42). 



With gooseberries and currants, however, the results appear 

 to be different one of the several instances in which these bushes 

 seem to differ materially from fruit trees (see pp. 29, 35, 107) 

 for the proportions were 



Half inch from ends. On rest of roots. On stems. 



57 38 5 



both kinds of bushes giving nearly identical values. 



Some further experiments were made to see whether the 

 omission of careful trimming of the old roots was detrimental 

 to the formation of new rootlets ; the trimmed and untrimmed 

 roots (purposely broken off in a rough manner) being equal in 

 number and diameter on each tree. Out of three series of 

 experiments, one showed no difference, either as regards the 

 total weight or stoutness of the new roots, and the other two 

 showed differences in opposite directions, so that the general 

 conclusion must be that the trimming of roots is a work of 

 supererogation. Here again, however, remarkable results were 

 obtained with gooseberries and currants, for they both showed 

 a great advantage from every point of view in favour of broken; 

 as against trimmed roots, the superiority applying to the rootlets 

 forming from all parts of the original root-system and the stems, 

 and amounting to an advantage of from 50 to 150 per cent. 

 This remarkable behaviour remains unexplained (cf. p. 35). 



ARRANGEMENT OF TREES IN A PLANTATION 



The simplest method of arranging any given number of trees 

 in a plantation, or, more strictly speaking, of arranging them 

 so that each tree has the same space allotted to it for, when 

 the area of the plantation is limited, the actual number of trees 

 which can be fitted in must depend to a certain extent on the 

 distances selected, and on the contour of the ground is on 

 what may be termed the square system (Fig. 8, A), with all the 

 trees equidistant from their neighbours in two directions at 



