446 Stoward. On Endospermic Respiration in Certain Seeds. 
In order to show how marked is the difference in the results obtained 
with sterile and non -sterile material, the following experiment with pure 
endosperms, which is contemporaneous with Experiment 28, and is an 
example of one that failed, may be adduced. The method of preparation 
was precisely the same as in Experiment 28. 
Experiment. Material prepared precisely as in Experiment 38. Steep 
water at termination of steeping found to be contaminated with micro- 
organisms. 
No. of 
objects . 
Wt. gnus. 
Vol. c.c. 
Temp. C.° 
After hours 
of respiration. 
^2 %• 
0*%. 
Mgms. Co 2 . 
COt/O,,. 
15 
1.9764 
3*25 
22° 
8 A 
n -75 
6.017 
0*04132 
0-84 
3 -i 5 
23 ° 
26J 
i 8'55 
0-00 
0.02135 
As comparison shows, during the first interval the output of C 0 2 in 
the above experiment amounts to nearly ten times that furnished by the 
pure endosperms in Experiment 28 conducted under sterile conditions. 
21 . Experiments to check possible Errors due to 
Apparatus. 1 
For the purpose of checking any error due to the Bonnier-Mangin 
(Aubert, 12) apparatus in the analyses of the respiration gases, the gas 
evolved in one of the experiments with Hordeum was transferred to another 
tube, and three separate analyses of it were made. 
The results are embodied in the following table : — 
Table. Check Analyses of Respiration Gas. 
Vol. of 
gas. c.c. 
Temp. C°. 
C 02 %. 
<?2 %• 
<^2 % <>2 % 
reduced to o° C. 
3-5 
27° 
7.79 1 
I 3 , 4 I 
7.090 
12-20 
3*25 
28 ° 
7-771 
13-43 
7.048 
12-lS 
2.90 
28 ° 
7*891 
13-38 
mean = 
7*157 
7.098 
12-13 
12.17 
1 Preliminary experiments with this apparatus, by means of which analyses of the air of the 
Experiment House were made, brought to light a source of error due to the mercury then used being 
contaminated with a small quantity of zinc, which readily oxidizes in atmospheric air. Thus the 
volume of air drawn into the calibrated tube of the apparatus underwent a diminution which could 
not be attributed to slight changes of pressure or temperature. When the mercury was subsequently 
shaken up with a solution of mercurous nitrate acidified with nitric acid and then washed with 
water and dried by filtering through a dry filter, this difficulty disappeared. 
All the experiments with Hordeum and Ricinus were carried out with the original calibrated tube 
supplied by the makers of the apparatus. This, prior to the experiments with Zea, was damaged, 
and had to be replaced by a locally made tube. On examination of this tube, it was found that 
there was a constant minus error in the determination of the C0 2 percentage of o-86 %, and 
consequently in all the C0 2 analyses in the experiments with Zea this amount has been added. 
Apparently the error was due to the bulb of the calibrated tube being irregular in shape, thus causing 
a small quantity of caustic potash solution to be persistently retained in it. 
