FLAVOR AND BENZENE HEXACHLORIDE CONTENT OF PEANUTS 15 



With one exception, there were no significant differences among 

 scores given by Panel A to the different commercial samples. The 

 score of 2.8 for one sample of peanut butter from Virginia Bunch 

 peanuts that followed cotton treated in 1950 with 80 pounds per acre 

 of BHC dust was significantly lower than scores for all other com- 

 mercial samples. There were more instances where judges reported 

 off -flavors described as characteristic of contamination by BHC in 

 this sample than in any of the other samples. However, various 

 comments, such as "sour/ 5 or "very objectionable off -flavors," suggest 

 that factors in growing or curing other than BHC contamination 

 may have been responsible for the extremely unpalatable flavor of 

 the sample. 



Palatability scores on the 1951 experiment station samples from 

 Holland, Va., are presented in table 5. Panel A rated 2 of the samples 

 from peanuts that followed cotton receiving 3.8 pounds of gamma 

 BHC per acre, and all of those from peanuts that followed cotton 

 treated with 5.1 pounds per acre, as significantly off -flavored in com- 

 parison with the control. While there is some overlapping of scores 

 between treatments, the general mean scores of 8.1 and 7.1 for the 

 samples following the low- and high-BHC cotton treatments, re- 

 spectively, are both significantly lower than the score of 9.8 given 

 the control sample, and also differ significantly from each other. 

 The results indicate that in this instance definite off-flavors in peanut 

 butter resulted from the growing of peanuts in soil on which BHC- 

 treated cotton was grown the preceding year. 



It will be seen also that among the 1951 experiment station samples 

 from Holland, Va., correlation of BHC content with cotton treatment 

 is generally comparable with that shown by panel scores (table 5). 

 Panel scores did not correlate perfectly, however, with the chemically 

 determined BHC levels. The fact that the control sample was gen- 

 erally scored 10, with a panel average of 9.8 out of a possible maximum 

 of 10, adds confidence to an assumption of real detection of off -flavors 

 in the samples from peanuts that followed BHC-treated cotton. This 

 assumption finds further support in judges' descriptions of observed 

 off -flavors. Among the 32 judgments on the samples from peanuts 

 that followed cotton given the 3.8-pounds-per-acre treatment, there 

 were 11 notations (35 percent of judgments) of off -flavors typical of 

 BHC. Among the samples from peanuts that followed the more 

 heavily treated cotton (5.1 pounds per acre), 18, or 56 percent, of the 

 scores were accompanied by such descriptions. No typical BHC 

 flavors were ascribed to the control samples. 



Panel B scores (table 5) show the same general tendency as those 

 by Panel A, though agreement on low-scoring samples was not perfect. 



Mean flavor scores for the 1952 samples from South Carolina, which 

 followed cotton treated with BHC containing 36 percent of gamma, 

 are given in table 6 and those that are significantly low are designated. 

 The ranked-pair analyses (p. 6) are not presented, but significant 

 findings were the same for both this method and the analysis of 

 variance. Results of the ranked-pair analyses also indicated that the 

 judges were scoring consistently with each other. 



Only 1 of the 4 treated lots from the Pee Dee Experiment Station 

 was significantly inferior to its control. The sample in plot B from a 

 field receiving 2 pounds of the gamma isomer of BHC per acre was 

 scored significantly lower than the samples from the untreated field 



