540. ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
perceived in the first medium that there had been an absorption of 
oxygen and a throwing off of carbon dioxide, and in the second 
medium asphyxiation. From this they concluded that respiration 
takes place in latent life and when it is not possible, the organism 
perishes; consequently, life in the embryo can be only relaxed. These 
conclusions allayed certain doubts as to the early experiments of 
Doyére and Davaine. This is why Lance in 1896 took up the study, 
confining his researches to arctisca. 
Contrary to the claims of the committee presided over by Broca, he 
himself affirmed that the coming to hfe of these beings is not a resur- 
rection: 
The arctisca of the roofs adapted to desiccation lose their power to revive when 
after desiccation they have been plunged into a gas, unsuited to support life, such as 
carbonic and sulphuric. When they find respiration impossible, they die; their 
latent life is then a relaxed life. 
We have therefore to deal with two contradictory hypotheses, 
apparently based upon facts equally conclusive: 
Is the relaxed life a more exact conception of the nature of latent 
life than the suspended life ? 
Must the one completely exclude the other or is each one partly 
true? These are questions which I have tried to elucidate and to 
which we shall now turn our attention. 
Ill. THE IMPERMEABILITY OF THE TEGUMENT OF CERTAIN SEEDS. 
When in 1904 I undertook these researches, limiting myself entirely 
to the latent life of seeds,! I asked myself if the prevailing contra- 
dictory opinions were not due to errors in interpretation of certain 
experimental results. For instance, were the embryos of the seeds 
really in contact with the media tried—confined air, irrespirable 
atmosphere, nitrogen, carbon dioxide, mercury, alcohol, chloroform, 
and ether? If their tegument had been impermeable, might it not 
have protected them against the various media that it was intended 
to subject them to? That was an important point, to which the 
greater part of my predecessors paid too little attention. 
It was therefore very necessary to find a means for determining the 
permeability of the teguments of the seeds which were most used in 
the above-mentioned experiments. I employed a very simple 
apparatus: A barometric tube closed at one end by a portion of the 
tegument to be experimented upon, then filled with mercury with all 
the usual precautions, and inverted in a dish of mercury. The varia- 
tions of the level of the mercury of this pseudo-barometer which 
terminated in a vegetable membrane, compared with the variations 
in the level of the mercury of another tube of the same kind, pre- 
1 Researches on the latent life of seeds (1904-1907). Annals of Natural Sciences, Botanical, 9th series. 
