November 24, 1905.] 



SCIENCE. 



679 



various soils upon growth much, more clearly 

 than do cultures in ordinary pots, vphere most 

 of the roots come to lie betvpeen the soil sur- 

 face and that of the pot. 



The soil of the present illustration is of the 

 Orangeburg clay type, from South Carolina. 

 Another sample of the same type from Texas 

 is used for comparison. The soil is generally 

 very fertile in both states. To investigate 

 the effects of fertilizers upon the sample from 

 South Carolina, wheat cultures in paraffined 

 wire baskets were grown for three weeks in 

 this soil with various treatments. By this 

 means the surprising discovery was made that 

 the untreated soil gave an exceedingly poor 

 growth of the wheat plants, and that it was 

 not appreciably benefited by any oue of a 

 large number of treatments used. An experi- 

 ment was then carried out to compare this 

 soil with the .Texas sample of the same type 

 and with Takoma lawn soil (see Bulletin No. 

 23 of this bureau), one of the poorest soils 

 with which the bureau has dealt. Transpira- 

 tion data from thirty plants for twelve days 

 showed that, considering the transpiration on 

 Takoma soil as 100, that on the South Caro- 

 lina sample was 41, while that on the Texas 

 sample was 209. On the same basis the green 

 weights of the plant tops grown on the two 

 soils were 35 and 216, respectively. Thus by 

 the basket method it appears that this par- 

 ticular sample from South Carolina is ex- 

 ceedingly unproductive, and that the unpro- 

 ductivity is not corrected by fertilizers. 



It seemed possible that the observed sterility 

 might be due to the presence of toxic organic 

 substances in the soil. Such toxic substances 

 have been found to be readily transmitted to 

 the soil extract and to show their character- 

 istic effects therein, and so to determine 

 whether or not such substances were present 

 here wheat seedlings were grown in aqueous 

 extracts of the two soils from South Carolina 

 and Texas. These extracts were prepared as 

 described in Bulletin No. 23, by stirring for 

 three minutes five parts by weight of soil with 

 six parts of water, allowing the mixture to 

 stand twenty minutes and then filtering ofE 

 the extract through a Pasteur-Chamberland 



filter tube. The plants of the present experi- 

 ment showed somewhat better growth in the 

 Texas extract, but the difference was not 

 marked. The transpirations from the two 

 cultures of sixteen seedlings each, grown six- 

 teen days, were, for the South Carolina 

 sample 118.4 g. and llY.l g., and for the Texas 

 sample 148.8 g. and 129.3 g., respectively. 

 This indicates that the soil to be tested con- 

 tains no soluble organic substances markedly 

 toxic to the plants and that it does contain 

 sufficient soluble inorganic material for nor- 

 mal plant growth. Its infertility must, there- 

 fore, be due either to some physical property 

 of the soil or to too great a concentration of 

 the soluble salts in the soil. The effect of 

 such concentration might be overcome to a 

 very large extent by the much greater dilution 

 of the soil extract. 



To determine whether the infertility might 

 be due to too great concentration of soluble 

 matter, the South Carolina sample was leached 

 by passing about an equal volume of water 

 through it, and the leached sample was com- 

 pared with the Takoma soil by another basket 

 experiment which ran for six days. Consid- 

 ering the transpiration and green weight each 

 as 100 for the Takoma soil, the figures ob- 

 tained from the leached soil were 196 and 152, 

 respectively. Thus by leaching the soil its 

 power to support plant growth was increased 

 from 41 to 196 by transpiration and from 35 

 to 152 by green weight. On the same basis, 

 the fertility of the Texas sample is repre- 

 sented by 209 for transpiration and 216 for 

 green weight, so that by leaching the sterile 

 sample it has been improved in fertility so 

 as to be nearly equal to the Texas sample. 

 These results show that the sterility of the 

 South Carolina sample, at least as far as 

 seedling wheat plants are concerned, is prob- 

 ably due to an excess of soluble salts. Chem- 

 ical analysis of the water extract of the South 

 Carolina sample showed the following amounts 

 of dissolved materials, expressed in parts per 

 million of the air dry soil by weight : NO3, 

 611; P0„ trace; SO,, trace; K, 100; Ca, 11; 

 and CI, 175. The large amounts of nitrate, 

 chloride and potassium here found seem to cor- 



