FLAVOR AND BEXZEXE HEXACHLORIDE CONTEXT OF PEANUTS O 



indicating no off -flavor; 4, perceptible off-flavor; 3, slightly strong 

 off -flavor; 2, moderately strong off-flavor; and 1, very strong off-flavor. 



A known reference sample was served at each testing session to 

 provide a uniform basis for scoring. Experimental samples to be 

 scored as "unknown" were matched with the reference as to variety 

 and degree of roast (estimated by color and flavor), in order to limit 

 the effect of these two variables on the comparisons. Three experi- 

 mental samples selected at random, were scored at each testing 

 session. Samples were scored at 3 -minute intervals to reduce fatigue 

 and allow for detection of aftertaste which might affect scoring of sub- 

 sequent samples. Judges were asked not to smoke or eat for 1 horn- 

 after each session and to report any "delayed perception" of off- 

 flavor if such occurred so that any samples with such off-flavor might 

 be retested individually. No definite evidence of delayed effects was 

 observed during these tests. In general, each commercial sample 

 was scored once b}^ the panel. Controls were scored as unknowns as 

 often as feasible, and the panel also rescored a limited number of 

 samples that were scored low the first time they were rated. Scores 

 for samples that were tested at a session in which the control was 

 included as an unknown were analyzed separately. 



Prior to the palatibility evaluation of the 1951 samples, additional 

 testing and training was given the judges. Available for this purpose 

 was peanut .butter that had been prepared earlier from peanuts grown 

 in soil treated with technical BHC at a dosage rate to give 1 pound 

 of gamma isomer per acre. Analysis of this sample showed a tech- 

 nical BHC content of approximately 7 p. p. m.; in preliminary tests 

 even untrained judges rated it as having very strong off-flavors char- 

 acteristic of BHC. Aliquots of this peanut butter were mixed with 

 a natural flavored peanut butter to provide test samples containing 

 approximately 3.5, 1.8, 0.9, 0.4, and 0.2 p. p. m. of BHC. Portions 

 of these samples were served, 3 at each panel session, to 14 prospec- 

 tive judges in a series of tests providing for decreasing amounts ol 

 off-flavor as the threshold test progressed. For this training a 10- 

 point rating scale was used with 10 indicating no off-flavor; 8, barely 

 preceptible off -flavor; 6, perceptible off -flavor; 4, slightly strong off- 

 flavor; and 2, strong off -flavor. 



The results of these training tests for prospective judges have been 

 recorded in an earlier publication {12). It will be sufficient to note 

 here that the tests provided for selection, from the 14 prospective 

 judges, of 6 individuals plus 2 alternates to serve on the panel for 

 judging the 1951 and 1952 samples. These selected judges all demon- 

 strated ability to identify the samples of peanut butter containing 

 1.8 p. p. m. of BHC and generally showed some indication of detec- 

 tion at the 0.9-p. p. m. level. Definite off-flavors were not observed 

 by any of the judges in the samples containing 0.4 and 0.2 p. p. m. 

 of BHC. 



Panel tests on the 1951 and 1952 samples were carried out as in 

 1950, the only difference being that ratings were made on the 10-point 

 scale rather than on the 5 -point scale used in 1950. 



With the 1951 samples, separate palatability studies were made 

 on the group of commercial samples and those from the Tidewater 

 Experiment Station. For the latter, the design used was a twice- 

 replicated balanced lattice with a randomized distribution of samples 

 for testing by 6 judges on 4 days. The reference sample was pre- 



