IT is well known that genuine proteins (at pH 7) are attacked by crystalline trypsin which, on the other hand, is able to split synthetic peptides. This has been taken as support for the view that these proteins contain peptide bonds in their molecules1. We wish, however, to point out that in the light of the following consideration, this support loses a great deal, if not all, its importance. If, according to Anson and Mirsky, denaturation is reversible, then in a solution of a given globular protein there is an equilibrium between genuine and denatured protein, G&rlharrD Hence it is sufficient that D and only D should contain peptide bonds open to fission by trypsin, because by removal of D by hydrolysis this process is forced in the direction from left to right and G will gradually disappear as well. The problem is open to experimental test in two different ways.