127 



The solution soon became cloudy and within a few 

 days a test for vanillin usino" the acid nitrate of mercury 

 reagent described by Estes (3) showed that the yanillin 

 had disappeared. The cloudy solution was plated out 

 and from the isolations made, bacteria were obtained 

 which, in pure culture, destroyed yanillin. 



By the kindness of Dr. (x. L. Fraps four of the soils 

 used b}' him in his work with yanillin were collected 

 and shipped in sterile cans. The four soils were those 

 referred to in his publication (4) as Nos. 876, 870, 1956 

 and 114. All four of the soils contained yanillin de- 

 stroying bacteria. 



From Dr. F. W. Upson the Black Meadow soil and 

 Lancaster fine sandy loam referred to in the work bv 

 Upson and Powell cited aboye were receiyed. These 

 soils were also collected and shipped under sterile con- 

 ditions. The presence of yanillin-destroying bacteria 

 in these two soils was also demonstrated and pure 

 cultures of yanillin-destroying bacteria isolated from 

 them. With these soils there was also forwarded a can 

 of what appeared to be quartz sand in regard to which 

 Dr. Upson stated. 



"We get a yery marked difference between the sand 

 and the two soils in regard to their abilit}' to destroy 

 yanillin and cumarin." 



In this sand no organisms destroying yanillin could 

 be demonstrated. 



Through the kindness of Dr. Oswald Schreiner and 

 Dr. J. J. Skinner, a sample of the Arlington soil was 

 receiyed. This sample was collected and shipped un- 

 der sterile conditions. Vanillin-destroying bacteria 

 were found to be present in this soil and were isolated 

 in pure culture. 



We haye, therefore, demonstrated the presence of 

 yanillin-destroying bacteria in those soils to which the 

 addition of yanillin has been found to haye little bad 

 effect on the growth of plants. The sand is of much 

 interest as here we apparently haye a case in which 

 the yanillin persists and evidences its toxicity because 

 of the absence of yanillin-destroying bacteria. The 

 Arlington soil is also of much interest because in this 

 case the vanillin persists and is toxic even though 

 vanillin-destroying bacteria are present m the soil. 

 We, therefore, assume that conditions are not suitable 

 in this soil for the growth and the action of the yanillin- 

 destroying bacteria present. To test the correctness of 



