Priestley and Ever shed. — Growth Studies . I. 227 
Generally ten determinations were made on each occasion, and the probable 
errors estimated (Wood ( 12 )). 
We are aware that these experiments are on a relatively small scale, 
but circumstances make it impossible for us to extend their scope appre- 
ciably. However, with regard to the experimental results and their value 
as a basis for new conclusions, it must be noted that the significant feature 
of the curves appears in every set of figures and in every case is associated 
with the same phenomenon. This characteristic feature remains when full 
allowance is made for the calculated probable error of the data, at the critical 
points of the curves. 
We hope that later further investigations, with increased experimental 
facilities, will permit of our conclusion receiving critical revision. 
3. Experimental Results. 
A summary of the results of our experiments will be found in the 
following tables and curves (Figs. 1-4). Brief details as to the con- 
ditions under which each set of growth data were obtained are given. The 
most significant figures are given by the Tradescantia experiments, and 
reference to the tables shows that this material has yielded much more uni- 
form results and that the experimental error is relatively small. 
Data obtained from the experiments with tomatoes, however, are valu- 
able in that they show, as the most striking features in a somewhat erratic 
curve (Fig. 4), the same characteristic points as in the case of Trades- 
cantia. 
Experiments with Tradescantia. 
Series I. The cuttings were started on June 5, 1920, and were taken 
from plants in the greenhouse at Weetwood Hall and grown in empty flower- 
pots which stood inverted in dishes of water. The cuttings selected were 
as uniform as possible, their weights varying between 2-5 and 2*8 grammes, 
and the number of leaves on each varying from 5 to 8. Ten cuttings were 
placed in each pot, and there were 20 pots in all. At intervals, one shoot 
was taken from each pot, and the roots were removed with a sharp knife, 
dried on blotting-paper, then weighed, dried at ioo° C. and weighed again. 
No roots were visible on the cuttings until the night of Thursday, June 10, 
and the first estimations of root-weight were made on June 14. Subse- 
quently the remaining cuttings were taken at various dates, but the last 
twenty not until July 5. 
The dry weights and wet weights of the twenty cuttings were recorded 
individually, and then the mean weight and probable error calculated. 
In the following table, only the mean weights and probable errors are 
given from consideration of space ; for the same reason, temperature and 
humidity records of the greenhouses are not given, though they are available, 
