IN a recent communication1 Goldstein suggests that differential staining of DNA and RNA can be obtained by a number of binary mixtures of basic dyes, of which the dye with the higher molecular weight always should stain DNA and the other one RNA. From the facts presented, however, this can only be taken as a working hypothesis for further investigations, since no evidence of a chemical nature is included. The author obviously takes it for granted that if staining is obtained in a morphological structure which is known or supposed to contain a certain chemical substance, this staining would be proof of specificity of the staining procedure for that substance. Of course this cannot be true. No cell constituent ever consists of pure nucleic acid; nucleic acids in the cell occur combined with basic proteins, whereas the structures in which these are found may contain many other substances, such as structural proteins, carbohydrates, lipids, etc. So, although specificity for a morphological entity may sometimes be claimed, this is still quite different from specificity for demonstration of a chemical compound. That is a main reason why only a very few of the numerous staining procedures of histology and cytology (many of which have been known for long to give colour differentiation between the nucleolus and the main nuclear material) have been recognized as reagents to specific types of substances, if properly applied2–4. In order to prove Goldstein's suggestion that differentiation between DNA and RNA should mainly depend on the relative weights of the cations of the two combined basic dyes, the following experimental evidence is required: