The data obtained in the series of SSCR experiments and the theoretical analysis of the results demonstrate the spectral shift control principle clearly. The critical assemblies show that considerably larger amounts of fuel can be loaded into a core moderated with a mixture of D/sub 2/O and H/sub 2/O than in one moderated with H/sub 2/O alone. The measured Cd ratios and the calculated neutron balances show that the excess reactivity associated with this extra mass is held up in the resonance absorption of the fertile material. This leads to a higher conversion ratio system than the same sized lattice controlled by thermal neutron poison. Very good agreement is obtained between experimental and theoretical values for nearly all important parameters over the entire range of D/ sub 2/O concentration in the moderator. This agreement confirms the use of the BPG-four-group treatment, and demonstrates the calculation of reactor criticality and other measured integral parameters from fundamental cross sections without the use of intermediate fitting functions. The reactor leakage part of the calculational model is shown to be accurate by the good agreement obtained between calculated and experimental values of age, criticality, change of criticality with moderator height, and the comparison betweenmore » boric acid poisoning and reactivity holddown with D/sub 2/O. Comparison of measured and calculated Cd ratios indicates that the model gives good values of resonance absorption for U/sup 238/ and of resonance fissibn for U/sup 235/. The accurate criticality results also confirm this. More cross section and lattice Cd ratio data are needed for Th before defini tive conclusions can be reached. In two areas theory and experiment do not sgree. Further analytical and experimental work is needed to obtain better results in the analysis of thermal flux ratios. There is definite disagreement between calculated and measured temperature defect, and effort should be exerted to resolve the discrepancy. The good agreement between theory and experiment justifies the use of the model in this series of experiments, but the experiments were all performed with the same moderator volume fraction and fuel. Further confirmation would be useful at other moderator volume fractions and fuel enrichments. (auth) « less