838 GEOKGE E. MISTOT AND AELIE V. BOCK 



those of groups 2 and 4. For a more detailed discussion regarding the is 

 agglutinins, iso-hemolysins and the selection of donors, the reader is 

 f erred to the references cited above and to those by Brem, Minot(&), Goes 

 Vincent (&), Sanford, Rous and Turner, Karsner(fr), Karsner and Koecl 

 ert, Clough and Richter. 



It is not the purpose of this article to disctiss technic, but it seems 

 desirable briefly to summarize a suitable method for performing these 

 agglutination tests. This summary is essentially the same as that previ- 

 ously given by Minot and Lee. 



In order to make a test between serum (fresh or stock) and the red 

 cells, the following simple procedure with chemically clean glassware 

 will usually suffice. A suspension of cells (about 5 per cent) is obtaine 

 by the addition of 3 to 5 drops of blood to about 2 c.c. of 1 per cent solu- 

 tion of sodium citrate in 0.9 per cent sodium chlorid solution. These 

 cells need not be washed. A drop of the red cell suspension is mixed 

 with a drop of serum. It is important to make the mixture complete. 

 This may be done upon a glass slide with a cover glass put over the 

 mixture. The cover glass should always be raised and the cells and serum 

 remixed several times before a negative reading is made. A hanging drop 

 preparation permits neater technic and avoids drying. The test often 

 may be read macroscopically, but should always be read microscopically, 

 in order to avoid any possible errors except when it is rapidly and un- 

 doubtedly positive. In order to guard against possible errors, it is always 

 wise to allow the mixture of cells and serum to remain for at least 30 

 minutes, preferable in the incubator. While there are few opportunities 

 for confusion in this simple test, nevertheless the penalty of transfusion 

 of incompatible blood may be so great that every reasonable care should 

 be given to the performance of the test. Confusion may be caused by 

 weak agglutination. It is always possible by employing different amounts 

 of cells and serum, by incubating the mixture for some hours and by 

 thoroughly washing the red cells, to decide the problem of doubtful reac- 

 tions. However, if by the method described the reaction is not cleai 

 and perfectly definite, the test must be repeated and perhaps amplified. 

 A safe rule is never to regard a reaction in which there is any doubt as 

 negative. Rouleaux formation may be easily demonstrated as quit 

 different from agglutination. Confusion may be caused by atypical agglu- 

 tinations, that are very rarely intense, due to auto-agglutination and allit 

 phenomena which are little understood. Stock sera for determining tc 

 what group a human being belongs will keep many months and even yt 

 if sterile, carefully sealed and in the ice-box. Stock sera have an ad- 

 vantage over fresh sera in that they are less liable than fresh sera tc 

 produce reactions with red cells, which may be confused with iso-agglu- 

 tination. It may be again emphasized that when carefully done the re 

 action of agglutination is in a very large proportion of cases clear anc 



