376 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



done in small quantities, at a large number of stations, with very 

 simple appliances. In view of the amount of seed which must be 

 handled each year, the cost of extraction can be materially reduced 

 and seed of higher average fertility obtained by concentrating the 

 major part of this work at central plants equipped with improved 

 machiner}^ Three extracting plants of tliis character are in process 

 of construction, on the Oregon National Forest, in Oregon, the Medi- 

 cine Bow National Forest, in Wyoming, and the Harney National 

 Forest, in South Dakota. 



The seed collected by the Forest Service shows a wide variation in 

 cost. This is due primarily to great dilFerences in the seed crops from 

 year to year. The cost of yellow pine seed collected in Montana has 

 ranged from 61 cents per pound in favorable years to $2.01 in seasons 

 of deficient seed crops, and tlie cost of Douglas fir seed from 84 cents 

 to S2.58. Since direct seeding, particularly under the broadcast 

 method, requires relatively large quantities of seed, the initial cost 

 of the seed bears a very important relation to the total cost per acre 

 of the completed work. 



A problem of still greater importance from the standpoint of final 

 results is that of having seed available at the season of the year when 

 needed for most effective use. Past experiments have demonstrated 

 that fall sowing is essential to success in most of the localities where 

 extensive seeding projects will be conducted. Experience has also 

 slio\vn that seed on a large scale can not be extracted in time for use 

 the same season. 



The purchase of clean seed from local collectors is often advisable 

 in localities where this industry has been developed and the quality 

 of the seed is assured. Such purchases must usually be made, however, 

 at prices above the cost involved when cones can be obtained by con- 

 tract or hired labor. Nor do such purchases meet either the problem 

 of excessive cost in seasons of short crops or that of inability to secure 

 the seed from year to year at the right time for successful sowing. 



European seed is exceptionally cheap. It does not vary greatly in 

 cost and can usually be obtained in the quantities and at the seasons 

 desired. In order to sow the acreage desired last year the collection 

 of native seed was supplemented by extensive purchases of Scotch 

 and Austrian pine at 43 cents per pound, Norway spruce at 32 cents, 

 and European larch at 76 cents. The low average cost of the seed 

 purchased as compared with that collected is due primarily to the 

 mclusion of these foreign orders. 



The results obtained in a large majority of the experiments con- 

 ducted up to the present, however, do not warrant further extensive 

 purchases of European seed. The adaptability of these foreign species 

 to soil and chmatic conditions in the western United States is too 

 questionable to justify their further use on a large scale until success 

 with them is assured by conclusive experiments. In the present stage 

 of this work it will be far preferable to rely mainly upon the valuable 

 native trees produced naturally under the very conditions with which 

 we are dealing, and to limit the use of exotics to small experiments 

 designed to show how far they may wisely be employed. 



Experience has demonstrated also that the quality of purchased 

 seed, both domestic and foreign, is much less dependable and subject 

 to control than that of seed collected by the Forest Service. This 



