23 



the season as determined by counts on different dates on 20 plants on 

 each plat. The important results of these counts are indicated in the 

 following table: 



Table VI. — Showing rate of fruiting on respective plats of Department of Agriculture's 

 bollworm experimental farm at Pittsburg, Tex., 1004- 



SUMMARY REMABK3 CONCERNING THE EXPERIMENTS WITH 

 CULTURAL METHODS. 



In the case of field experiments involving questions of the character 

 of those presented above, final conclusions ma}" not be drawn as the 

 result of one or even several years' tests. Variations in seasons and 

 other conditions often produce results one 3'ear not verifiable the next; 

 and what appears desirable for one character of soil may not be use- 

 ful on another. The writers Avould, therefore, be entirely unwar- 

 ranted in making specific recommendations, especialh" as regards the 

 use of certain fertilizer elements, on the data secured in the course of 

 last year's investigation. 



Certain general statements are, however, apparentl}" warranted. At 

 the Pittsburg, Wills Point, and other experimental farms the use of 

 fertilizers resulted in a notably earlier and larger crop of cotton, as 

 compared with the unfertilized check plats. 



On soils deficient in one or more of the three principal elements of 

 plant food, namel}", phosphoric acid, nitrogen, and potash, the appli- 

 cation of the needed element or elements results in a more rapid and 



