344 



SCIENCE. 



[N. S. Vol. XIX. No. 478. 



Wliile it. is thus explicitly stated, and it is 

 a matter so notorious as to admit of no ques- 

 tion, that crop yields are often affected ad- 

 vantageously by proper fertilizers, it is main- 

 tained that such substances can not be held 

 as alone the chief factor in determining yield 

 of crop, since climate, soil management, etc., 

 produce effects of the same order of magnitude . 

 as do the fertilizers, and that it may happen 

 that the several effects wotild nullify one an- 

 other in any particular season, illustrations 

 almost innumerable being on record. 



Attention may also be called to the fact that 

 the bulletin does not attempt to treat specifi- 

 cally of commercial fertilizers, nor of their 

 use in practice, but the matter is brought into 

 the text only as a necessary consequence of the 

 discussion of the crop-producing power of 

 soils. No claim to an exhaustive presentation 

 of this subject was made. 



It is also maintained, and the reasons there- 

 for clearly stated, that no scheme of chemical 

 analysis yet proposed can, in itself, determine 

 the fertility or crop-producing power of a soil. 

 A chemical procedure is described, novel in 

 some respects, which the authors of the bul- 

 letin used in their researches, but it is made 

 so evident as to allow of no misconception that 

 this procedure has proved and would generally 

 prove as futile as all its predecessors in attempt- 

 ing to show the probable productive capacity of 

 a soil or its fertility. This is not the place to 

 enter into a discussion of the technical rea- 

 sons for the inadequacy of our analytical pro- 

 cedures to measure or estimate fertility, but it 

 is safe to say that the position taken, in regard 

 to this point at least, is in full harmony with 

 that of the best authorities.* To cite two re- 

 cent utterances on this point, at the meeting 

 of the Association of Agricultural Colleges 

 and Experiment Station Officers held in Wash- 

 ington last November (1903) Director Thorne, 

 of the Ohio Experiment Station, in describing 

 the results of plot experiments extending over 



* Prom the many citations which could be 

 given the following is taken as one of the most 

 conservative: Bailey (Cornell University Agr. 

 Exp. Sta. Bull. No. 119, 1896) states, 'a chemical 

 analysis of soil is only one of several means of 

 determining the value of land, and in the general 

 run of cases it is of secondary value.' 



a number of years, stated that it was difficult 

 to see how the results could possibly have been 

 anticipated by laboratory examinations of the 

 soils. At this same meeting Dr. H. W. Wiley, 

 chief of the Bureau of Chemistry, U. S. De- 

 partment of Agriculture, stated : " When a 

 man sends to me a specimen of a given soil 

 and writes, 'Please analyze this soil and tell 

 me what crops I can grow on it,' I send him 

 word, ' Ask your soil itself what you can grow 

 on it; in that way asking your question di- 

 rectly of the soil, you can get your answer, 

 and in no other way.' " At a later point in 

 this address it was explicitly stated that if 

 chemical methods could be devised for deter- 

 mining the food constituents in soils, different 

 procedures must of necessity be devised for 

 extracting each constituent from the soil, and 

 different procedures again for each crop. 



Hopkins delivered an address at the meeting 

 in Washington already mentioned, and has 

 anticipated the publication of the proceedings, 

 the address having appeared as Circular No. 

 72, Agricultural Experiment Station, Uni- 

 versity of Illinois. In it exceptions are taken 

 to Bulletin No. 22, partly through evident 

 misinterpretation of the text; partly through 

 disapproval of the use which the authors have 

 made of the well-known data from the Eotham- 

 sted Station, although the validity of the 

 conclusions drawn is in general admitted; and 

 partly because it has been possible on the basis 

 of chemical analysis, to advise the use of fer- 

 tilizers containing potassium on certain Illi- 

 nois soils, with improved yield of crop. The 

 relevancy of this last argument is not ap- 

 parent unless it is meant to imply that the 

 same method of analysis would always lead 

 to as favorable results, a conclusion unfortu- 

 nately disproved by numerous instances on 

 record. Indeed, it is a matter worthy of 

 notice in passing that such an instance is cited, 

 without explanation, on page 10 of Circular 

 No. Y2 of the Illinois Experiment Station. A 

 soil containing according to analysis an enor- 

 mous amount of nitrogen (6Y,000 pounds per 

 acre), an abundant amount of phosphorous 

 (2,000 pounds per acre) but what is regarded 

 as a deficient amount of potassium (1,200 

 pounds per acre) produced no corn when either 



