1. Three modifications are suggested in the KOESSLER and HANKE colorimetric technique (based on the Pauli reaction) for the estimation of pyrimidine bases: (a) the substitution of bicarbonate for some of the carbonate in one of the reagents (leading to greater stability in the pigment formed); (b) the use of carnosine nitrate as a standard of reference for estimation instead of a mixture of dyestuffs; and (c) the use of an appropriate light filter in conjunction with the colorimeter. This leads to an eightfold increase in sensitivity.2. Extracts from symmetrical pairs of muscles from the legs of the frog give Pauli reactions of equal intensity. This equality is not destroyed by previous fatiguing or killing by heat of one of the pair.3. The cutaneo‐crural, sartorius, gracilis, semitendinosus, semimembranosus, and gastrocnemius muscles contain increasing concentrations of carnosine (measured by the Pauli reaction) in the order named.4. Different frogs from one batch differ considerably in the apparent carnosine content of any one muscle; variations in the gastrocnemius, for example, were observed from 124 to 293 mg. per 100 grm.5. The Pauli reaction given by histidine, histamine, carnosine, and extracts of living and dead muscles have been studied. In the conditions chosen, all save histamine were alike in the tint given and in the rate of colour production and fading.6. No substance capable of giving the Pauli reaction was found to diffuse from living muscles into a surrounding saline whether the muscles were resting or fatigued, and regardless of the presence or absence of oxygen.7. Histidine can diffuse from saline into an excised resting muscle and out again. The equilibrium reached appears to be a simple osmotic equilibrium.8. All substances capable of giving the Pauli reaction diffuse freely out of a muscle in rigor (induced by heat or by iodoacetic acid).9. Carnosine has been isolated in the form of its copper salt in comparable amounts both from muscles killed while in the resting state and from those previously sent into heat rigor.10. Carnosine diffuses freely and completely into a dead muscle, but if the muscle is alive only 30 per cent. of its bulk appears to be permeable to carnosine diffusing in.11. It is concluded that the carnosine in living muscles of the frog is contained within some series of structures occupying only 70 per cent. of the bulk of the muscle. It is retained in virtue of some property of the tissue which is lost at death.12. No evidence was found for the existence either in dead or living muscle of substances other than carnosine capable of giving the Pauli reaction.This work has been financed by the Medical Research Council.