The selection of new anode-cathode combinations for primary cells is described on the basis of ampere-hour capacity, theoretical and operating potential, handling properties with respect to the design of primary batteries, and limitations placed on the selection of active components by unusual electrolyte requirements. Theoretically, for the design of dry and reserve primary cells, the most desirable anode materials are magnesium and aluminum; the most desirable cathode materials are the nitro-organic compounds, N-halogen organic compounds, silver (II) oxide, copper (II) oxide, nickel oxide, and mercuric oxide. Such other possible anode materials as titanium, alkali metals, and hydrocarbons, and cathode materials such as oxygen, sulfur, halogens, oxides, and chlorides, are also considered. A magnesium/magnesium perchlorate/m-dinitrobenzene dry cell with an experimental capacity of 90 wh/lb and estimated maximum available capacity of 150 wh/lb is described. The basic relationships between the cathode potential of aromatic nitro compounds and the nature of substitution groups are illustrated. The development of high-capacity dry cells with a perchlorate electrolyte, a magnesium anode, and a cupric oxide or synthetic manganese dioxide cathode is described. The capacity of a magnesium/magnesium perchlorate/mercuric oxide cell on a 30-minute discharge rate with a voltage tolerance of ±5 per cent is 50 wh/lb and 3.0 wh/cubic in.