Apvams—Effect of very low Temperatures on Moist Seeds. 3) 
_ 12 per cent. of moisture. But the dry seeds are not frozen, or, at 
any rate, are not killed. Consequently we must look for some 
other explanation. 
The intercellular spaces in seeds are small; and after freezing 
has gone on for some time in moist seeds, the intercellular spaces 
will become a solid mass of ice. If the temperature is still further 
lowered, then it seems more likely that complete rupture takes 
place, causing a separation of the cells from each other, or it is 
conceivable that ice crystals may be formed inside the cell within 
the protoplasm itself, and death of the protoplasm may be due to 
mechanical effects. But whatever the explanation of what takes 
place, it seems fairly certain that freezing to death can only occur 
if the seed contains more than 12 per cent. of moisture. 
Davenport, in his “ Experimental Morphology,” summing up 
the results of experiments on dry seeds, spores, &c., concludes 
with the generalisation, “‘ That there is no fatal minimum 
temperature for dry protoplasm.” It would, I think, be justifiable 
to extend this generalisation, and to maintain that there is a fatal 
minimum temperature for moist protoplasm, and that this fatal 
minimum lies somewhere above the temperature of liquid air. 
But it will be at once obvious that this latter generalisation is 
apparently contradicted by the experiments of Macfadyen’ with 
Bacteria. He kept a number of species in broth cultures immersed 
in liquid air for seven days, and a second series in liquid 
hydrogen for ten hours, and found that they still retained their 
vitality. It must be observed, however, that these experiments 
were qualitative and not quantitative. In all probability what 
actually happened was that all the Bacteria in the actively 
vegetative condition were killed, that some were in the spore 
condition, and as their protoplasm was in the dry state, these 
survived and reproduced rapidly when the unfavourable conditions 
were removed. Klepzofi’s’ results with Anthrax were entirely 
different from Macfadyen’s. 
More recently Smith and Swingle® carried out more than 
1 Proc. Roy. Soc., vol. lxvi., 1900. 
2 Centralbl. f. Bakteriol. u. Parasitenk. 1te Abt. xvil., 1895. 
3 Science, 31 March, 1905. 
