543 The method has been applied to hydrolysates of human y-globulin and of a protein from the surface antigen of PEeudomna aerugino8a L 2 (Mead & van den Ende, 1953) and also to concentrated normal urine. The globulin hydrolysate gave spots corresponding to aspartic and glutamic acids, cystine, serine, threonine, tyrosine, glycine, alanine, valine, leucine (isoleucine), phenylalanine, proline, histidine, lysine and arginine. Spots at the positions taken by 13 of the common amino acids were detectable on the ionophoretogram given by the antigen-derived protein. The urine concentrate served to indicate that the method may be applied directly to solutions rich in dissolved solids. Taurine, glycine, serine, asparagine, methionine, alanine and histidine were tentatively identified. Two-dimensional ionophoresis should give better separation than one-dimensional in all cases where change of pH affects the relative as well as the absolute rate of migration of the substances to be analysed. In practice, however, difficulties in finding sufficiently volatile buffer components may be a limiting factor. Mixtures of proteins are an obvious field for experiment also lipoprotein-polysaccharide complexes such as some bacterial antigens for at least one of which (that of P. aerugino8a strain L 2) the excellent method of detection on paper due to Rydon & Smith (1952) may be used. SUMMARY 1. In a new apparatus for two-dimensional paper ionophoresis, the pH of the buffer solution in the paper is changed, after ionophoresis in one dimension , by exposure to the vapour from a solution exerting a suitable and substantially constant partial pressure of ammonia. 2. A method for determining the flow of electro-lyte at different points on the paper is proposed. The results of applying it in the new apparatus are discussed. 3. The following amino acids are separated qualitatively by the new apparatus: glutamic and arginine. Leucine and isoleucine are not separated. The developed ionophoretogram is ready within nine hours of starting an experiment. I am grateful to Professor M. van den Ende for encouragement and to Dr A. Polson for many helpful discussions. The apparatus was constructed by Mr W. H. Borret without whose skilful workmanship this investigation could not have been carried out. In recent years, chromatographic procedures using ion-exchange resins have been applied to the separation of a large number of substances, but their successful application to the separation of proteins has been confined mainly to basic proteins of low molecular weight, such as cytochrome c (Paleus & …