r3THE significance of the relationship of size to failure of commercial banks lies in its bearing upon proposals for 1 strengthening the banking structure. If, as is generally thought, the small bank has contributed disproportionately to the failure list, proposals looking to increased minimum capital requirements are urgently needed. On the other hand, if bank mortality is made up in about equal proportions of all sizes of banks, other remedies should be sought. The present article is concerned entirely with the relation of size to failure. It is limited to national banks, first, because failures among this class have been sufficiently numerous to give representative results, and second because the data are readily available and reported according to uniform methods. National banks in operation on September 8, x920, as reported by the Comptroller of the Currency, were divided into ten groups on the basis of total resources. National banks failing during the twelve-year period from July I, I920, to June 30, I932, were divided on the same basis.2 The use of I920 as a base, though simpler than using the failures each year of each size of bank as a percentage of the banks of that size in existence at the beginning of that year, involves certain difficulties. It is possible that the proportions of banks by size groups were considerably different in I932 from those in I920. Thus the failure of large numbers of small banks in certain geographical sections may have reduced consid-