172 On the Vitality of Farm Seeds. 



^vitllout killing the seed, so long as the moisture is not removed. 

 Air-diied seeds (among which were oats and peas) have been 

 exposed for four days to a temperature of 330" to 345" Fahr. 

 below freezing point, and thereafter slowly thawed ; their 

 germinative power showed no decrease from this treatment. 

 There is an upper limit of temperature, somewhere from 130" 

 to 200" Fahr. (or, with special precautions, even higher), at 

 which chemical changes taking place in the live tissues cause 

 their death. The temperature in my cabinet never, of course, 

 came anywliere near this limit, consequently the gradual death 

 of the seeds in it was due not to chemical alteration produced 

 by temperature, but to the steady loss of moisture going on 

 continually at ordinary air temperatures. Once the seed is dead, 

 other changes of a very gradual character go on, due to slow 

 oxidation, and in the end give the whole seed a charred appear- 

 ance, like that produced by burning, i.e., by rapid oxidation. 

 Such are the geniiine grains of '' mummy wheat," and the 

 grains found in the Romano-British granaries hollowed in the 

 chalk of the south-west of England. While no seed could 

 possibh' survive under the conditions in which mummy wheat 

 and the Roman granary wheat were placed, it is possible for a 

 thick-coated seed under favoural)le conditions to live a very long 

 time. Seeds of the sacred Lotus of Egypt (Neluml)o) ger- 

 minated after having been for a hundred j-ears in a cabinet in 

 the British Musetim, and for many years before that in Sir 

 Hans Sloane's keeping. The duration of life of different seeds 

 is no doubt chietiy determined by the amount of protection 

 afforded the embryo by its coverings. 



Rapidity of Germination. 



The details are given in the tables appended to this paper. 



In these experiments records were also kept to ascertain the 

 rapiditj^ of germination, a character in which seeds differ 

 greatly. Looking at the tables here given, we see that in l(SiM), 

 the first year of the experiments, no less than eleven kinds of 

 seed completed their germination within a week, viz., barley, 

 white oats, meadow fescue, timothy, white clover, sainfoin, the 

 two swedes, the two turnips, and rape. At the other extreme 

 are the species where not more than half the living seeds ger- 

 minated within the first week, such as sheep's fescue (50 per 

 cent.), sweet vernal (49 per cent.), cocksfoot (45 per cent.), 

 wood meadow grass (37 per cent.), and smooth-stalked meadow 

 grass (5 per cent.). 



As to the effect of age on rapidity of germination, one would 

 naturally expect that the older seeds would need time to regain 

 the portion of natural moisture lost by keeping before absorbing 

 the further moisture necessary for germination. The time 



