1046 JOURNAL OF FORESTRY 



remained for six years unaffected, and when placed under suitable con- 

 ditions germinated as quickly and apparently with as much energy as 

 fresh seed. The inference is that the more resistant seed of this species 

 may lie in the ground many years before germination takes place. 



These experiments by Davis are interesting, in that they are in accord 

 in a general way with the results obtained by foresters in the study of 

 the germination of tree seeds. Evidence is rapidly accumulating to the 

 effect that some of the seeds of many species of forest trees lie on or 

 in the ground for long periods before germination can take place. It is 

 possible that in some cases at least this is due to the resistance which 

 the seed-coats offer to the intake of water. 



J. W. T. 



The Botanical Gazette. Vol. LXIV, August, 1917, pp. 166-167. 



The questions centering around the distribution 

 Natural Rcpro- and frequency of seed trees has been subject to 

 duction from much discussion in recent years in the United 

 Seed States. A few years ago J. V. Hofmann began 



Stored in a study in District 6, United States Forest Serv- 



the Forest Floor ice, to determine the efficiency of seed trees in 

 restocking and the distances from them that seed 

 is disseminated. The results of this study and the investigation of 

 reproduction, following extensive burns in the Puget Sound region, led 

 to the investigation of the viability of seed stored in the duff and the 

 surface soil of the forest, and the possibility of its being the source of 

 a widely and evenly distributed reproduction following extensive lum- 

 bering operations and fires rather than the scattered seed trees left in 

 lumbering or on the margins of burns. 



From comprehensive studies on the average number of seedlings per 

 acre at various distances from seed trees, both on burned and unburned 

 slash, it appeared that coniferous seed trees are effective for a distance 

 of only two or three chains, and that reproduction occurring in un- 

 burned slash is distributed without relation to the distance from seed 

 trees. 



The above conclusions led to the investigation of the litter and surface 

 soil in order to ascertain if they contained germinable coniferous seed 

 in quantity to account for reproduction. Hofmann reports large num- 

 bers of seed in the duff, which germinate as soon as the stand is felled 

 or moderately burned, thus admitting light and heat to the forest floor. 

 The series of studies upon which his conclusions are based extended 



