586 SCIEXCE PROGRESS. 



shaft. No GlaiiciiLvi had previously grown in the neighbour- 

 hood. White (3) suggests " that they (the seeds) had 

 possibly remained inert from the time when the deposition 

 of the Lias took place, and upon their exposure to the 

 atmosphere were recalled to life". — This is not only ex- 

 travagant, but transcends imagination. 



An interesting case of reputed 'longevity,' that lacks 

 however positive proof, is one recorded by v. Heldreich 

 in 1873. A heap of mine-refuse on Mt. Laurion in Attica, 

 3 m, deep, had been removed shortly before he visited the 

 spot. On the recently exposed surface a Glaucitnn w^as 

 found, which he did not hesitate to consider as a new 

 species, G. Serpieri ; Silene Juvenalis Del., hitherto un- 

 known there, was also found. To account for their appear- 

 ance, he suggested that they had been dormant under the 

 refuse-heap since it was first placed there, at least 1 500 

 years ago. De Candolle (2) quotes a statement of A. P. 

 de Candolle, who germinated seeds of Mimosa pudica after 

 these had been dormant sixty years, and one of Girardin, 

 who succeeded with seeds of Phascoliis vulgaris (haricot), 

 that had lain more than 100 years in the herbarium of 

 Tournefort. It is further stated that in 1850 Brown sowed 

 seeds from the collection of Sir Hans Sloane, of which they 

 had formed part longer than i 50 years, and caused them to 

 germinate. The latter cases seem quite trustw^orthy ; one 

 might however object that the seeds had possibly been 

 changed unknown to the experimenters. 



Experiments of Peter (4) support the view that the 

 sudden appearance of species hitherto unknown, when soil 

 has been disturbed, is due to persistence of seeds of one 

 vegetation dormant in the earth while later vegetations suc- 

 ceed and to their subsequent germination when conditions 

 changed. He tried to find if soil holds seeds, if it be able to 

 preserve their ' vitality,' as well as which species have seeds 

 capable of remaining thus unputrefied. Soils were tested 

 on the surfaces of which vegetation had not existed for a 

 long time ; present covering with wood was disregarded, 

 since tree-seedlings admit of easy distinction and could be 

 eliminated. Soils were chosen concerning which it was 



