Expts were done to yield information concerning the relative backscattering powers of substs. at 2 diff. detection geometries. A 4 microg. sample of Cl4-active BaCO3 was mounted over an area of 0.04 sq. cm. in the center of a plastic film circle 20 cm. in diam. and 0.07 mg./sq. cm. thick; the sample layer was not thicker than 0.15 mg./sq. cm. The Al equivalent thickness of the counter window and air path was 3.4 mg./sq. cm. at the lower geometry (12%) and 2.3 mg./sq. cm. at the higher (36%). The sample was counted over 25 cm. of air; thick layers of materials were placed within 0.05 mm. of the back of the sample spot and the activity measured. This activity divided by the first equals the backscattering factor of the subst. in the backing layer at the geometric efficiency with which detection was carried out. Considering radioactive BaCO3 mounted on Al and counted at 36% geometry, if the sample is made of many thin layers, the observed activity of the 1st lamina is 1.16 times the intrinsic activity since the Al contributes radiation flux by exterior reflection. The activity from the next lamina is increased by more than 1.16 since the radiations are more powerfully reflected from the first lamina. As thickness increases, activity rises from 1.16 to 1.35 times that when all reflection effects are neglected. Backscattering effects saturate rapidly because they involve double transit of radiations through absorbing layers. The reflection effects reach 80% of their max. at a sample thickness of 6 mg./sq. cm. and 97% at 12 mg./sq. cm. The effective self-absorption corrections for several sample substs. are not very nearly the same as those for Ba CO3. At sample thicknesses where the backscattering effects are saturated, the curves for the 2 sample materials will be related by the quotient of the proper reflection coefficients. The value predicted is 1.35 [plus or minus] 0.01/1.07 0.015 = 1.26 0.02; that observed is 1.27.