Yagi, Georges Gilbert, and the crew of the 

 Venus during 1 day's fishing approximately 30 

 miles off EXamond Head in the vicinity of the 

 Penguin Banks . 



Absorptions of serums were performed 

 by mixing 1 to 2 ml of a suitably diluted ser- 

 um with 0. 1 ml. of packed erythrocytes in a 

 small centrifuge tube. The cells in these mix- 

 tures we:ce kept suspended by hand or machine 

 agitation for periods extending to 20 minutes at 

 room temperature. After this time the absorb- 

 ing cells were removed by centrifugation . One 

 absorption generally sufficed to remove the 

 antibodies for the antigens on the cells involved. 



EXPERIMENTS WITH INTACT 

 ERYTHROCYTES 



Individual variation in skipjack antigens. 

 Ten individual skipjack were found to fall into 

 four categories which were distinguishable by 

 variations in the specific affinities of their 

 erythrocytes for antibodies in various serums. 

 Fish of the first category (fish Nos. 11, 13) 

 carried an antigen (1) with a relatively strong 

 affinity for antibodies in various rabbit anti- 

 serums prepared against the blood of yellowfin 

 tuna, albacore, skipjack, and white croaKers. 

 Fish of the second category (fish Nos. 12, 14, 

 17, 18, 19) had an antigen (2) with a relatively 

 strong affinity for "natural" antibodies occur- 

 ring in certain "normal" serums (human, bo- 

 vine and sheep). Fish of the third category 

 (fish Nos. 10, 15) had both antigens 1 and 2 on 

 their cells. One fish (No. 16) of the fourth 

 category had neither of the antigens (1 and 2) 

 noted above. For convenience, the occurrence 

 of antigens 1 and 2 is shown in all the tables 

 that follow by superscripts following the num- 

 bers of individual fish. 



Table 1 shows that fish carrying antigen 

 1 (Nos. 10, 11, 13, 15) differ markedly from 

 those not carrying this antigen with respect to 

 the degree to which their cells were agglutin- 

 ated by selected rabbit antiserums. The 

 numbers in the line labeled totals were simply 

 obtained by adding the numerical estimates of 

 the individual reactions above and are pre- 

 sented as a convenient single number for 



comparative purposes. (The reactions of fish 

 No. 16 appear to involve an antigen other than 

 those described in the text as 1 and 2) . 



Table 2 shows the results of absorbing 

 1 in 50 dilutions of albacore No. 10 and oceanic 

 skipjack antiserums with cells of selected fish 

 (the technique of absorption is described in the 

 section on materials and methods). These re- 

 sults are in harmony with the concept that fish 

 Nos. 10, 11, 13, and 15 carry an antigen 1 that 

 distinguishes them from the other fish studied. 

 The white croaker and mackerel antiserums 

 shown in table 7 were also absorbed at 1 in 50 

 dilution using cells of Nos. 12 and 13. In both 

 cases the No. 12 cells left antibodies for No. 13 

 cells, while No. 13 removed antibodies for both 

 fish. Antigen 1 appears from these data to have 

 a greater ability to combine with the heterogen- 

 eous antibody populations used in its detection 

 than does antigen 2. This suggests a structural 

 relationship between the two antigens that is 

 similar to those classically described for the 

 A, and A„ subgroups of human A antigen. 



In addition to the several reactions 

 noted above, a 1 in 4 dilution of the normal ser- 

 um of fish No. 12 (with antigen 2), weakly but 

 definitely agglutinated the cells of fish Nos. 10, 

 11, 13, and 15. This reaction, shown in table 

 3, was confirmed by a duplicate test and by the 

 use of undilute 12 serum. The serum of fish 17 

 (also with antigen 2) gave, in the duplicate test, 

 an indication of agglutination with the cells of 

 fish Nos. 10 and 11, but the reaction was too 

 slight to be systematically studied. Although 

 alternative interpretations are also possible, 

 this observation is in agreement with the con- 

 cept that antigen 1 occurs on the cells of these 

 fish. 



Populations of antibodies with relatively 

 strong affinities for an antigen, 2, on the cells 

 of fish Nos. 10, 12, 14, 15, 17, 18, and 19 

 were discovered in the normal bovine, sheep, 

 and human blood grouping serums described in 

 the section on materials and methods. Table 4 

 shows the reactions of selected skipjack cells 

 with these serums. 



