LATENT LIFE—BECQUEREL. 543 
existence of a respiration in seeds. These results afford a proof that 
the gaseous exchanges demonstrated by Van Tieghem and Bonnier 
are not due to an attenuated respiration, but to a simple chemical 
oxidation of the surface of the tegument or of the embryo. 
The generally accepted conception of the latent life of seeds must 
be modified. It is an extremely sluggish, intracellular, anaérobic, or 
else a suspended life. How is one to choose between these two 
hypotheses ? 
V. LONGEVITY OF SEEDS. 
If the life of seeds in nature were entirely suspended, if all the pro- 
toplasmic functions of assimilation and of disassimilation were com- 
pletely arrested, as claimed by Claude Bernard, thew germinative 
power should be unlimited. This is what many naturalists believed 
when they were told of the extraordinary case of the longevity of 
grains of wheat inclosed for more than 2,000 years in the tombs of 
the Pharaohs, which, once sowed, would have germinated. But it is 
now known that the good faith of these scientists was imposed upon. 
Mixtures of authentic and recent grain were sold to them. This 
fraud, by which such botanists as Alphonse de Candolle and Decaisne 
were not deceived, was unmasked by M. Maspéro. This eminent 
egyptologist never succeeded in germinating the grains of wheat which 
he himself collected in the tombs of the Pharaohs. Furthermore, the 
study of these grains made by Ed. Gain showed that their embryos 
were partially destroyed; when they were moistened, they were trans- 
formed into an amorphous pulp. 
On the other hand, no confidence can be placed in the story of seeds 
from Roman sepulchers or the graneries of Cesar, Argau, or Hercu- 
laneum, or from Merovingian tombs or excavations. Too many flaws 
in the evidence, ignored by the investigators, destroy every basis for 
their claims. Only experiments made with specimens of which the 
time of harvesting the seeds and the date of their arrival in the 
laboratory are known can give us acceptable evidence. 
Already, in 1831, Alphonse de Candolle had carried on researches 
with 368 kinds of seeds preserved for 14 years in sacks. Many species 
of Leguminose and Malvacez had conserved their germinative fac- 
ulty. I resumed the work of that learned naturalist, extending it to 
500 kinds of seeds belonging to 30 of the more important families of 
monocotyledons and dicotyledons. The seeds came from the seed 
collection of the Muséum d’Histoire Naturelle of Paris. The time of 
their collection, carefully verified, varied between 25 and 135 years. 
Four families furnished germinations: the Leguminose, the Nelum- 
bonacex, the Malvacez, and the Labiate. 
Twenty of these germinations came from seeds 28 to 87 years old. 
Among the Leguminose the oldest species were Cassia bicapsularis 
