64 SCIENCE PROGRESS 



treatment. As seeds have been investigated from this point 

 of view, some points concerning their longevity may be con- 

 sidered here. Exposed to ordinary conditions, with free access 

 of air, the seeds of different species vary widely in their 

 length of life. Very short-lived are those of some kinds of 

 poplars and willows. In a set of trials cited by Wiesner, the 

 fresh seeds of Salix purpurea germinated within twenty-four 

 hours, whereas older seeds required from two to four days 

 to initiate germination : of these seeds kept for forty days 

 21 per cent, germinated, kept for sixty days only 10 per cent., 

 and kept for eighty-five days only 8 per cent. The results of 

 the two tests of vitality of the seeds, namely time required for 

 the inception of germination and the percentage of those 

 capable of germinating, together suggest that death is no 

 sudden catastrophe but is a gradual process of weakening. 

 Before attempting to explain this gradual collapse, additional 

 evidence pointing in the same direction may be quoted. In 

 Dobner and Nobbe's Botanik fur Forstmanner there are 

 given statistics as to the percentages of seeds germinating 

 after these had been kept for periods varying from half a 

 year up to twelve years. The species tested included Pinus 

 sylvestris and Trifolium pratense. The results are given in the 

 form of percentages of germination in seeds of the same 

 sample but of different ages. If the statistics be arranged in 

 the form of curves, whose ordinates represent the percentages 

 and whose abscissae note the ages of the seeds, the resultant 

 "curve of mortality" is a descending one having a primary 

 steep descent succeeded (excepting in one sample of Spergula 

 arvensis) by a gradual shallow concluding fall. The regularity 

 of the descent is such as to indicate a steady influence at work 

 killing out the seeds (for I assume that the investigators 

 eliminated the possible source of error by determining the 

 penetration of water through the testa). The curves printed 

 here are those obtained from different samples of seed of Pinus 

 sylvestris. 



What is the gradual lethal influence ? Some interesting 

 experiments conducted by Giglioli (see Nature, 1895, p. 54) 

 provide the first steps towards solution of the problem. Giglioli 

 kept resting seeds in hermetically sealed tubes that contained 

 various dry gases, including such poisons as chlorine, arseni- 

 uretted and sulphuretted hydrogen and carbon monoxide, also 



