This analysis of the scattered radiation was made with double x-ray spectrometer previously described. (Davis and Purks, Proc. Nat. Acad. Vol. 13, No. 6, June 1927, Vol. 14, No. 2, Feb. 1928.) A special double target (Mo) x-ray tube was used with a block of graphite between the targets as the scattering element. The crystals were placed for reflection at first order. Four lines were observed: (a) an undisplaced line at position of Mo $K{\ensuremath{\alpha}}_{1}$; (b) a line displaced 80\ensuremath{''} (0.0012A) from $K{\ensuremath{\alpha}}_{1}$ position; (c) a line displaced 140\ensuremath{''} (0.002A) from $K{\ensuremath{\alpha}}_{1}$; (d) a line displaced 780\ensuremath{''} (0.0113A) from $K{\ensuremath{\alpha}}_{1}$. These are all on the long wave-length side of Mo $K{\ensuremath{\alpha}}_{1}$. The undisplaced line (a) is scattered from whole atom. The line displaced (0.0113A) closely agrees with $h{\ensuremath{\nu}}^{\ensuremath{'}}=h\ensuremath{\nu}\ensuremath{-}Ve$, where $\mathrm{Ve}$ is the energy level (287 volts) of the carbon atom. The line (b) is displaced (0.0012A). This is equivalent to 29 volts which does not agree with the ${L}_{111}$ level of carbon (11.2 volts). The line (c) is displaced (0.002A). This is equivalent to 50 volts, and does not agree with the ${L}_{1}$ level of carbon which is 34 volts. In the case of these two lines the relation $h{\ensuremath{\nu}}^{\ensuremath{'}}=h\ensuremath{\nu}\ensuremath{-}Ve$ does not hold. A search was made for scattered radiation on the short wave-length side of the Mo $K{\ensuremath{\alpha}}_{1}$ position, corresponding to the relation $h{\ensuremath{\nu}}^{\ensuremath{'}}=h\ensuremath{\nu}+Ve$. No positive results were obtained.