Viral nucleic acids have been found to be infectious for tissues and animals, yet are nonantigenic and resistant to antibodies against whole virus. Other unique properties, such as their host range and their susceptibility to nuclease action, render them a wholly new dimension to be reckoned with in virology. These properties may explain a number of conditions which at present are considered anomalous. The release from infected tissues of even a small proportion of total virus as free nucleic acid could, in an otherwise immune individual, lead to a low level of infection which would, perhaps, explain permanent immunity. If the proportion of nucleic acid is higher, or if the nucleic acid is resistant to nucleases because of an inert envelope, such conditions as "carrier" states-that is, viremias with or without antibodies-are possible. Note added in proof: The recent article by J. D. Ebert and F. H. Wilt (54) came to my attention after the present article was written. Ebert and Wilt's excellent article indicates the impact of the newer knowledge of viral nucleic acids on the ideas developing in the field of embryology.