579 
Balls . — Temperature and Growth . 
extreme case of the latter is seen in the cells of the cotton seedling, where 
the volume of the culture medium is so small that staling takes place in 
a few hours at so low a temperature as 30° C. 1 
Dilution of stale media. 
If the idea of increasing concentration which has been sketched above be 
true, we ought to be able to simulate increasing staleness by adding the 
stale media to fresh media. 
Before doing so we have to eliminate the possibility that the ratio of 
< x 1 to the available food might have some subsidiary effect. This was 
done by four parallel experiments, using a solution which was on the verge 
of complete staleness : — 
50 % Stale ; 50 % Fresh ; stopping-point, 32*1° C. 
50 % „ 50 % Water ; „ „ 32-5° C. 
50% Fresh; 50% „ „ „ 37-1° C. 
100% „ ..... „ „ 37-5° C. 
Hence, if any such effect is exerted it is inconspicuous beside the effects of 
adding * x\ 
The first experiments gave erratic results, on account of the slow 
destruction of i x } by boiling 2 , and owing to the time wasted by this the 
results of the dilution experiments are not of a quantitative nature at 
present. The transfer of stale media from one tube to another is now 
effected through a flame, thus avoiding the necessity for subsequent 
sterilization. 
When the stale medium (or ‘ solution of # ’) is added to a fresh medium, 
we observe a depression of the stopping-point which appears to correspond 
to the percentage of ‘ x ’ which has been added. 3 The lowest depression 
obtained with a 50 % mixture of perfectly stale medium with fresh medium 
was 8° C., the stopping-point being 29-5° C. 
Parallel with this is the fact that when such graded solutions are 
inoculated and allowed to stand at a constant temperature, the rate of 
growth is proportional, as far as macroscopic evidence can be trusted, to the 
percentage of ‘ x * which is present, 3 provided that the experiment is not 
carried on so long as to involve secondary staling of the media. 
I have not yet been able to show that a proportionate relation exists 
between the depression of the stopping-point and the amount of * x * which 
has been added to the medium. That such a connexion exists is almost 
beyond doubt, but until its exact nature shall have been determined 
I consider that the present research is of more interest than value. When 
we have correlated the influences of the various proportions of ‘ x’ with the 
1 See p. 558. 2 Page 581. 
3 Note also the three growth -gradients in Curve 2. 
R r 2 
