The red blood cells of Mr. R. W., an 86‐year‐old type O patient with Proteus mirabilis wound infection, anemia, and thrombocytopenia were found to be polyagglutinable due to Tn‐activation, to be agglutinated by and to absorb anti‐A, and to have a reduced sialic acid level probably as a result of enzyme modification in vivo. The reaction with anti‐A was inhibited by blood group A substance. It is possible that both Tn‐activation and acquisition of A‐like antigen were changes induced by Proteus infection but, if so, the changes persisted after the bacterial infection was cured. The red blood cell modification in this patient has features in common with others in whom T‐ or Tn‐activation, reduced sialic acid levels, or acquired B antigens have been found. Various abnormalities of the red blood cell membrane are discussed and clinical, serological, and physicochemical findings are compared.

Red cells were treated with 20 different trypsin preparations. When the red cells were used for detecting Rh-antibodies, the trypsins gave comparable results provided the solutions had the same trypsin activity as measured by Anson units. When testing for warm haemolysins, the effect on the red cells was dependent on the activity as measured by Anson units of the single trypsin preparation, but there were great differences between trypsin preparations which indicates some kind of qualitative differences between these latter. The findings also suggest that there are differences in specificity between warm haemolysins from different people.

A "new" antibody, named anti-Joa, defining a high-frequency erythrocyte antigen has been found in the sera of two Negro patients. The two bloods are mutually compatible but a large number of other erythrocyte samples lacking common antigens are Jo(a+). A sibling of one patient is Jo(a+) but no other family studies were possible.Anti-Joa reacted with 3,000 group 0 but otherwise unselected blood samples, mostly from Caucasians, and with 768 blood samples from Negroes.