86 
Journal of Agricultural Research 
Vol. XXI, No. I 
Table II. —Mean growth increments oj shoots and concentration of sap of apricot trees 
(1918) 
Date. 
Mean length. 
Increment of 
growth dur¬ 
ing preceding 
week. 
Osmotic pres* 
sure of cell 
sap. 
May 2 
Cm. 
45- 0 
Cm. 
19. O 
Atmos¬ 
pheres. 
12. 71 
8 
60. 6 
& 18.2 
11. 68 
15 
74- 1 
*3- 5 
' n-99 
21 
«3-9 
6 II. 4 
10. 87 
29 
92.4 
? 7-4 
ii- 95 
June 7 
103.9 
& 9. O 
12. 30 
12 
k>S-9 
& 2. 8 
14. 96 
19 
108. 5 
2. 6 
i3- 5 2 
26 
113. 0 
45 
13. 86 
July 3 
122. 6 
68.4 
13. 66 
IO 
134-2 
11. 6 
14. 24 
17 
143-3 
9.1 
16. 05 
24 
148. 0 
4-7 
15.28 
3 1 
I 55* 1 
7-1 
12. 47 
Aug. 7 
162. 2 
7- 1 
14. 87 
14 
167. 4 
5-2 
14. 88 
21 
170. 0 
2. 6 
i5- !7 
28 
173-9 
_ 3-9 
15- 28 
Sept. 5 
« 184. 2 
6 9. 1 
15- 54 
12 
189.3 
5- 1 
16. 42 
19 
192.3 
3-o 
13.46 
26 
196.7 
4.4 
15- 30 
Oct. 3 
I 99- 3 
2.6 
15.64 
9 
201. 0 
b 2. 0 
14.52 
16 
203. 2 
2 . 2 
14. 99 
23 
204. 6 
i- 4 
18. 36 
3 1 
206. 8 
6 1. 9 
18. 90 
Nov. 6 
207.3 
•5 
17. 66 
a Interpolated value. & Calculated to 7-day basis. 
The degree of association between the amount of growth in a week and 
the concentration of the cell sap at the end of that week is more succinctly 
expressed by the coefficient of correlation, which was found to be 
r = —0.613 ±0.079. 
This coefficient is of sufficient magnitude to express a strong negative 
correlation between growth and sap concentration. The regression of 
the two variables is approximately linear (fig. 2), though there is a 
marked tendency for the points to scatter. These determinations are 
based on 28 observations, however, and are regarded as somewhat more 
reliable than those for the walnut trees. 
These observations, made at frequent intervals throughout a growing 
season, show that rapid growth is marked by a lower sap concentration 
and vice versa. In material of this kind, the concentration of the cell 
sap is probably due mostly to sugars and other organic compounds. 
