POSITION OF BRANCHES. 



119 



The means and standard deviations for these distributions are given in 

 table 63. It is seen from these tables that: 



(a) In mean position successive branches are roughly one node apart. 

 This is of course the result Vv^e should expect to get if a branch were 

 formed at every node. But while the latter is far from being the case 

 in detail, the branches are so distributed as to give an average result 

 of much the same kind. The regularity of the increase in mean position 

 is shown in fig. 24. 



(b) Neither the mean nor the modal position of the branches falls 

 at the node we should expect if a branch occurred at each node. Thus 

 first branches occur on the average at the third node, instead of the 

 first, fifth branches at the seventh or eighth node instead of the fifth, 

 and so on. In the case of the Series IV plants the mean position of the 

 branches is quite uniformly two nodes in advance. 



(c) Each of the first five branches shows about the same degree of 

 variation in its position. In the Series IV plants the variation apparently 

 decreases somewhat in the branches beyond the second, but the differ- 

 ences are so small that no stress can be laid on them. 



(d) The tendency for branches to occur at odd nodes with greater 

 frequency than at even, which was observed in the case of the first 

 branches, does not appear to hold for other branches, nor beyond the 

 third node. 



Branch 



Fig. 24.— Graphs of the means given in table 64. Series I, II, and III com- 

 bined, > ■ ■ n i l I ; Series IV, o — — — - . 



