488 Priestley and Pearsall. — Growth Studies. III. 
As the whole apparatus is in practice filled with liquid from a large 
reservoir at room temperature, the error introduced by temperature changes 
during the course of the reading is certainly well below other sources of 
experimental error. 
Experience shows that if the plant used is only wedged into position in 
the neck of the container by means of its cotyledons, it is liable to move 
as the cotyledons decrease in size. In future, therefore, it would be advisable 
to fix the plant upon a ledge in the neck of the apparatus. 
The accuracy of the results obtained with this apparatus can be 
estimated from the first series of readings made with it. Five containers 
were in use at the time, each containing a growing plant. After each 
reading, the container was emptied and the whole reading repeated. The 
average divergence of the second reading from the first was 5*50 ( x o-oi c.c.), 
the figures (unit = 001 c.c.) for the individual containers being as follows : 
No. of container 
1 
2 
3 
4 
5 
No. of observations 
26 
28 
26 
24 
26 
Total 130 
Average divergence 
Cn 
Oi 
00 
5^4 
5*39 
5.00 
5* s 9 
Mean 5-50 
Maximum divergence 
12 
15 
12 
19 
The largest divergence is 19 (x o-oi c.c.), and the usual character of the 
volume changes recorded (readings being taken every second day) can be 
gathered from the following summary of the distribution of the actual volume 
changes obtained within a few arbitrarily chosen limits : 
Increase in volume o-oi c.c. per 2 days Over 100 100-60 60-20 20-10 below 10 
v No. of observations 5 9 69 26 21 
This set of observations lasted considerably over a month and included 
not only readings when growth was practically complete, but also the 
temporary cessation of growth following on the development of lateral 
roots. During the normal development of bean roots, volume changes of 
the order of 0-2 to o-6 c.c. (per two days) are to be expected, after the 
appearance of secondary roots. The earlier stages of root growth are 
probably better observed by direct measurements of length as in Leitch’s 
experiments ( 1 ), or by the method of Neilson-Jones ( 2 ). 
One further point justifies our assumption that the volume changes 
observed lie outside the normal range of experimental error. The main 
results obtained are reproducible, the growth curves of individual plants 
being of the same general type. We hope to present conclusions based on 
these results in a subsequent paper. 
Bibliography. 
1. Leitch, I. : The Effect of Temperature on the Rate of Growth of Roots of Pisum sativum. 
Ann. Bot., xxx. 25, 1916. 
2. Neilson-Jones, W. ; A Simple Root Auxanometer. Ann. Bot., xxxiv. 555, 1920. 
