The basis for the difference between A1 and A2 specificity has been in dispute from the moment of its discovery (1-5). An excellent review of the earlier studies may be found in the thesis of Gammelgaard, recently available in English (6). One view holds that the A~ and A~ antigenic determinants are chemically identical, differences being essentially quantitative and ascribable to the presence of many more determinants on an A1 than on an A~ erythrocyte (7-11). Based on hemagglutination-inhibition titers of salivas from A~ and As secretors, B oettcher (11) proposed that individual molecules of A substance from A1 secretors have a higher ratio of A:H antigenic determinants than those from A~ secretors. He also proposed that the A~ gene is a duplication of the As gene. Bar-Shany et al. (12) found only a small, although significant, difference between A~ and A2 saliva in absorption studies. A proportion of A2 bloods contain anti-A~ (13-17). M~ikelii et al. (18) recently attempted to explain why anti-A~ does not agglutinate A2 cells without assuming that the A~ and A~ determinants differ. Hypothesizing the receptor density to be higher on A1 than on A2 erythrocytes, Miikel~i suggests that anti-A~ antibodies have a low affinity for A receptors and cannot form firm bridges between red cells using only one valence per cell. Thus they would only agglutinate A1 cells on which the receptors are close together. A~-specific antibodies therefore would have to be multivalent and of the 7M type and bivalent antibodies such as 7G would not distinguish between AI and A2 erythrocytes.The other view contends that there is a qualitative difference between the antigenic determinants in A1 and As individuals (19)(20)(21)(22)(23). Friedenreich and Worsaae (23) suggested that A~ and A~ are qualitatively different but related antigens. Their assumption was based mainly on absorption experiments showing that anti-A serum contained at least two kinds of antibodies, anti-A and anti-A~. A1 cells reacted strongly with anti-A