Periodical Literature. 337 



tire has for years been advanced by practitioners and theorists, 

 some assigning it to soil conditions, especially the formation of raw 

 humus under the change of light conditions, others to the inter- 

 ference of beech litter or other causes for unsuitable seed bed, 

 snow accumulations, weed growth, others again to the interfer- 

 ence by game, by fungus, by insects. 



Finally, in 1904, somewhat extensive trials throughout the 

 region were inaugurated to determine the cause of the failing re- 

 generation ; 25 groups on 15 trial plots were similarly located, 

 each group of 600 square feet. These were made in 3 series of 5 

 plots each, the first series was left unprotected to test the influence 

 of game, the second series in full light, the third in subdued light 

 which was secured either by the neighboring stand or by lath 

 screens. 



The first set was sown on the natural soil cover, the second set 

 after removal of the loose litter on the unworked humus, the third 

 with the humus layer worked in with the mineral soil, the fourth 

 with both litter and humus cover removed, the fifth with the 

 mineral soil worked after removal of litter and humus. 



The seed was gathered from the 200-year old firs on the ground, 

 and was sown as Nature does it, without cover, at the rate or 1 

 grain to the square meter (90 lbs. per acre). After four years, 

 during which the plats were under observation, the trials were 

 considered concluded. In all cases the loss of plants during the 

 four years was considerable, averaging /Q% of the original stand. 

 In the first series only 11% survived; in the second series 31% in 

 the average and 37% on the best, the last set; while the third 

 series (in subdued light) showed 19% survivors, varying from 12 

 to 21 on the various plots. 



Taking series 2, the most advantageous, by itself, and noting 

 the losses from year to year, it is noticeable, that the percentage of 

 loss from year to year decreased rapidly, the loss per cent, averag- 

 ing 42, 29, 18, 8 from the first to the fourth year, but, with the ex- 

 ception of the plots left in natural condition the proportion of loss 

 from plot to plot showed little difference, the last two series being 

 only slightly favored. 



On the plots left natural, at the end of the period there were 

 still over 5,000 seedlings per acre, which would be quite satis- 

 factory, but 8% of all the plots had no plants, 21% at best not 

 over 1,000, 8% at best not over 2,000 seedlings, so that altogether 



