16 _ BOTANICAL GAZETTE [JULY 
in the present investigation the same phenomenon was observed 
macroscopically, under after-ripening and germinating conditions. 
Such an increase during after-ripening is characteristic of seeds 
with dormant embryos. 
On the other hand, in Amaranthus retroflexus catalase activity 
appears to be far less subject to fluctuation. CrRocKER and Har- 
RINGTON (16) find surprisingly slight variation in the catalase 
activity during after-ripening, which in Amaranthus occurs “during 
the first three or four months in dry storage.” The activity for the 
first month and a half after harvesting, as shown in table III, is 
maintained at a very uniform rate. A comparison of the values 
obtained on the imbibed seeds with those found by Crocker and 
HarrINncTON for the dry powder indicates the uniformity of the 
degree of activity in different seeds: 
OXYGEN (CC.) LIBERATED AFTER 
WEIGHT OF POWDER (GM.) 
I minute 5 minutes ro minutes 
0.10 dry powder..... 4.9 9.0 bi Ae 
o.13 imbibed seeds. . . 3-9 6.1 x 
The values for the dry seeds, however, are slightly higher than for 
the imbibed, probably owing to the greater concentration of material 
in a given weight. In this connection the results obtained by 
CROCKER and HARRINGTON on samples of Amaranthus collected in 
1894 are of interest. They found the catalase activity of these 
twenty-three year old seeds but little diminished, although there 
was complete loss of viability. 
| OXYGEN (cc.) R AFTER ; PER CENT 
DATE OF COLLECTION GERMINATION 
: : AFTER 
| I minute | _ § minutes | ro minutes DAYS 
| 
1917 (average of 3 samples). 8.3 20.0 23.8 100 
1894 (average of 3 samples). 7.8 17.8 20.8 ° 
The extreme stability of the catalase activity is emphasized by the 
fact that one 1894 sample gave values identical with those obtained 
from one of the 1917 samples: 
