Barley plants were grown in complete culture solution and in deficiencies of phosphorus, nitrogen, or potassium for a period of about 6 weeks. Excised roots of these plants were treated with a complete, aerated culture solution at 25° C. for varying periods of time, and the changes in respiration rate, phosphorus, nitrogen, potassium, sugars, and starch contents measured. There were changes in fresh weight and dry weight of the excised roots during treatment. The dry weight decreased with time but the water-content changes were variable. There was a gain or loss of water by the roots according to the treatment. In all cases the deficient roots increased in content of the element in which they were originally deficient. The roots of the plants supplied with full nutrient usually decreased in content of phosphorus, nitrogen, and potassium, but exceptions occurred and the reasons are discussed. In most of the experiments described simultaneous loss of one ion and gain of another occurred. Nitrogen-deficient roots accumulated nitrate when exposed to a complete nutrient solution, and some of this was assimilated with formation of protein. Under similar conditions nitrogen-rich roots decreased in nitrogen content and proteolysis took place. There was a rapid fall in sucrose and reducing sugar content of the excise' roots. The starch content was initially very small and showed little change with time. The respiration rate declined with time in all treatments except where a nitrogen deficiency existed. Here the respiration rate increased to a maximum value at about 8 hours and then fell. This increase in rate is attributed to protein synthesis. No evidence of a ‘salt respiration’ was observed even when active uptake of phosphorus or potassium was occurring. In most instances the carbon dioxide evolved in respiration greatly exceeded the carbon dioxide equivalent of the sugar consumed in the same period. Exceptions were found with the nitrogen-deficient roots where less carbon dioxide was evolved than the equivalent of sugar consumed. It is probable that a part, at least, of the sugar unaccounted for was used in protein synthesis. Where the carbon dioxide of respiration was in excess of the equivalent of sugar consumed, protein or amino-acid is the most probable substrate. Respiration rate is found to be related both to nitrogen and sugar content.