Sept, i, 1931 
Growth in Branches of Young Pear Trees 
865 
with the growth of other laterals on the same mother shoot and that the 
correlation between adjacent laterals is greater than for nonadjacent 
laterals, suggesting that the factors determining the amount of growth 
are not uniformly distributed in the mother shoots. Moreover, the 
greatest correlation is found between laterals at the distal end of the 
mother shoots and becomes less with each successively lower pair of 
laterals. 
The coefficients of partial correlation go somewhat further in showing 
these relations. A comparison of r 12t r U2 , and r 124 shows that by equal¬ 
izing the effect of the third lateral, the coefficient is diminished from 0.346 
to 0.266, but equalizing the effect of the fourth lateral reduces it only from 
0.346 to 0.324. Broadly speaking, the length of the fourth lateral has 
less effect than that of the third upon the correlation between the length 
of the first and second laterals. When the length of both the third and 
fourth laterals is equalized the correlation between the first and second 
laterals is reduced from 0.346 to 0.258. The value of r u is the lowest of 
all the gross correlations and the coefficients of partial correlation into 
which r 14 enters are so low in comparison with their probable errors as 
to lack statistical significance. 
The values of these coefficients may, therefore, be regarded as good 
evidence in favor of an integral relationship in the growth processes of 
neighboring laterals on an upright mother shoot and that neighboring 
laterals have a tendency to vary together. It would seem difficult to 
escape the conclusion that factors which promote growth in any lateral 
also tend to promote growth in the adjacent lateral. It would at least 
seem that these measurements speak against any explanation which 
rests upon the assumption that the dominant apical shoot is such because 
it has withdrawn food or growth stimulators from the subapical region 
of the mother shoot. It is quite possible that two distinct groups of 
factors are operating to influence the growth of these laterals, one of 
which tends to make all the laterals of a given mother shoot either longer 
or shorter than the average for the population as a whole, while the 
other group tends to oppose the development of the subapical shoots, this 
suppression being the more complete as the distance from the distal end 
of the shoot increases. The existence of significant positive correlations 
between adjacent laterals speaks for the existence of the first group of 
factors. The systematic differences in the values of the interlateral 
correlation coefficients speak in favor of the second group of factors. 
This may mean that while there is a well-marked tendency for the 
factors which produce length in the first lateral to produce length in the 
laterals below it, there is some other factor at work which tends to block 
the operation of the length-producing factor in the subapical shoots and 
that it is more completely blocked as one proceeds down the axis of the 
mother shoot. 
