The only treatment by which nonspecific staining could be satisfactorily removed was by passing the conjugate through a DEAE-cellulose column. When 1 ml of conjugate was passed through a column (** f), the first and second milliliter fractions collected were the most specific and gave no nonspecific staining in some experiments, and very little in others. In the latter cases an additional treatment of the DEAE cellulose treated ** f with 50 mg of sweet clover stem tissue powder further improved the specificity. After these treatments the conjugate did not stain healthy or crown gall sweet clover tissues or stained them a very faint green which was easily distinguishable from the bright yellow-green specific staining. With this purified conjugate the best staining procedure consisted of treating the sections with 1: 10 dilution of ** f for 30 minutes, washing with PBS for 15 minutes, staining with ** f for 30 minutes, and washing with PBS for 15 minutes. The specificity of staining in WTV tumors with ** f and ** f but not with NS and ** f or with antiserum to potato yellow-dwarf virus and ** f, and the absence of such staining in crown gall tumor tissue from sweet-clover, indicate that an antigen of WTV was being stained.

** f was first conjugated with 50 mg of FITC per gram of globulin. This conjugate was passed twice through Dowex-2-chloride and treated with various tissue powders in the same manner as described for the indirect method. In all cases a disturbing amount of nonspecific staining was still present although it was still distinguishable from specific fluorescence.