252 M. A. de Candolle on Germination 
determining accurately, I immediately made a fresh sowing at a 
temperature of 24°-4 to 24°°9; it sprouted in twenty-two hours 
and a half. 
Sinapis appeared to have germinated in thirty-six hours ; but 
it was in the night, and the moment was not ascertained. 
Trifolium germinated about the forty-second hour. 
Nigella and Iberis escaped observation, from an accident. 
Lepidium presented a singular fact, probably resulting from an 
error of observation or the accidental choice of more tardy seeds 
than the others. This species, which germinates rapidly at low 
temperatures, only commenced partially to germinate (two 
grains out of ten) towards the end of the sixth day, and most of 
the seeds sprouted between the sixth and the seventh day. The 
temperature of the seven days varied from 22%] to 25° 1, the 
mean being about 23°°6 or 23°°7. The construction of the curve 
(Plate IV.) shows that this fact is not m harmony with those 
deduced from higher or lower temperatures, consequently that 
there was some error or accident. ‘To satisfy myself further, 
I repeated the experiment in November in another form, with a 
lamp placed under a large flask of water, im which a bottle 
containing a sowing of Lepidium floated. The mean temperature 
was 21°], with insignificant variations, and the Lepidium ger- 
minated after thirty-eight or thirty-nine hours, exactly as in the 
above-mentioned experiment. At a temperature of 26° to 27°, 
which unfortunately rose much higher (43°) during some hours, 
the Lepidium began to sprout at the sixteenth hour. Hence we 
may conclude that the experiment at 25° was inaccurate. 
Lastly, Collomia did not germinate in July. The temperature 
of the sand remained, from the 24th of July to the 3rd of August, 
between 22°°5 and 22° 1; on the 8th of August it fell to 18°5, 
and then rose on the 14th of August to 28°6. The seeds had 
been preserved and watered. I thought they would not germi- 
nate; but on the 15th of August two of them did so. The mean 
temperature varied too much for the experiment to be satisfactory. 
Assuming it to be accurate, it would be necessary to admit that, 
at a mean of 21°5, Collomia requires a period of twenty-seven 
days, which agrees moreover with the observation at 19°-6, as 
shown in the tracing of the curves in the Plate. It might be 
questioned, as in the case of Lepidium, whether the temperature 
of the second half of the period, which was momentarily lowered to 
18°5, had not caused the germination which the heat prevented 
in the preceding period. I doubt this, however, because the ger- 
mination took place when the mean had returned to 20°6. 
Moreover the duration of twenty-seven days agrees tolerably 
with that of the experiment at 17° to 18°, as is well shown by the 
curves. . 
