182 AlMd J. Ewart: 



The most complete set of soil records has been made by Peter 

 (I.e.), by notino; the appearance of plants in soil taken from 

 forests of known age, without undergrowth, and in which the soil 

 was supposed to have been undisturbed since the forest was 

 planted. He found certain plants, mainly meadow and field 

 weeds, continually appeared where a forest had covered meadow 

 or held for 20 tf) 40 years, but mainly sylvan plants from still 

 older forests, and hence concluded that the seeds in question 

 had lain dormant in the soil during- the life of the forest. This 

 evidence has been accepted even in Pfeffer's Physiology, although 

 the logic is childish in the extreme. In nearlj' every case, as can 

 be seen from the lists, the seeds in question have a short duraition 

 of life, rarely exceeding five years, but they are all either small 

 or i^eadily dispersed by wind, or animals, thus reaching the forest 

 land from neighbouring meadows, etc., and being covered more 

 or less deeply by the activity of burrowing animals, ants, etc., 

 or falling down cracks and burrows in the soil. If the matter 

 were so simple as Peter imagines it to be, and these seeds were 

 derived from the original vegetation prior to the forest, we should 

 expect to find them more abundant some distance below the 

 surface than on the latter itself. Those on the surface are. more 

 apt to germinate and decay, and Duvel hais shown that the more 

 deeply buried seeds last longer. In addition, the humus from 40 

 to 100 years of forestal growth would be sufficient to deeply 

 cover the original surface. On the other hand, if the soil is con- 

 tinually receiving fresh supplies of seed from wind aoid water, or 

 from the excrements, feet and bodies of birds and other animals, 

 some of which seeds are buried from time to time by burrowing 

 animals or washed down cracks, burrows or holes jn the ground, 

 we should naturally expect to find the surface richer in seeds 

 than the deeper layers. Peter found germinable seeds down to a 

 depth of 32 cms. (1 foot), but in only two cases were a few more 

 seeds found at 8-lG cms. (3-6 inches) than on the surface, and 

 the number 24 to 32 cms. deep was from 3 to 20 times less, and 

 never greater than in the surface layer. Peter's observations are 

 good evidence of the readiness of dif;persal of certain seeds, but 

 as evidence of their longevity are quite untrustworthy. They 

 contain iii grain of truth buried in a mass of inaccuracy. The 

 same applies to aill similar records of supposed old seed in soil 



