-u- 



In the second phase of the study, samples of infested x^heat having 

 test weights of 56, 58, 60, and 62 pounds per bushel, were tempered to 

 moisture contents of 10, 12, and ll; percent for each test weight. Sam- 

 ples of each lot were immersed in a 2-la7ered, nonmiscible combination 

 of a sodium silicate solution iri.th a specific gravity of I.I60, floating 

 on a methyl chloroform-deodorized kerosene solution vjith a specific 

 gravity of 1.30, A specific gravity of I.I60 of sodium silicate was 

 selected as one which irould give a reliable separation, and still allow 

 a broad margin of variability both in procedure and in wheat character- 

 istics. 



RESULTS 



By immersing samples in sodium silicate solutions ranging in speci- 

 fic gravity fl*om 1,0^0 to 1,3^0, varying degrees of separation of weevil- 

 infested from noninfested wheat were obtained (table 1). 



At the lower part of the range, about $0 to 70 percent of the in- 

 fested kernels were floated off the grain samples with no sound kernels 

 floating. At all specific gravities between 1,050 and 1,190 the float- 

 ing layer contained only infested kernels, and no sound kernels. The 

 proportion of infested kernels that appeared in the floating layer in- 

 creased until all infested ones floated at 1,350. However, at this 

 point, many lightweight sound kernels floated alsoj 56 percent of the 

 floating layer at 1.350 was composed of sound kernels. 



The degree of separation resulting from immersion in the 2-layered 

 combination of samples of wheat varying in weight and moisture content 

 is presented in table 2. The 1,300 specific gravity solution of methyl 

 chloroform and deodorized kerosene floated the kernels with early stages 

 of infestation which sank in the l,l60 specific gravity solution of 

 sodium silicate. 



There did not appear to be an outstanding relationship between the 

 degree of separation and the test weight or the moist^jre content. With 

 three exceptions, the top floating layer contained only infested ker- 

 nels. Nearly all of the floating kernels contained advanced stages of 

 larvae. The middle layer contained a mixture of kernels infested vjith 

 early stage larvae and lightweight sound kernels. The bottom layer x^as 

 composed mostly of sound kernels. Figure 1 presents a view of a t;ypical 

 separation. 



ANALYSIS OF RESULTS 



These studies indicate that an absolute separation by flotation of 

 infested from noninfested kernels probably will not be possible, but 

 flotation may be used to indicate the presence of hidden infestation 

 quite accurately. 



