'38 



NA TURE 



[December 2, 1922 



Human Blood Relationships. 



THE idea that a loss of blood by hemorrhage or 

 the possession of blood of a poor and deterior- 

 ated quality might best be rectified by the introduc- 

 tion into the body of blood from a healthy person is of 

 respectable antiquity. It is small wonder that the 

 ancients attributed to so splendid and conspicuous a 

 tissue an importance rather beyond its due. About 

 the time of the fire of London Pepys attended experi- 

 ments in which the blood of one dog was passed into 

 another and found to be sufficient for its needs, and on 

 another occasion at which a man was hired for a 

 sovereign to have some sheep's blood let into his body. 

 For even at this time it was realised that some sorts of 

 blood were more suitable for transfusion into man 

 than others. Little boys might be bled to death in 

 the fifteenth century to provide stimulating potions 

 for aged Popes, but human blood seems scarcely to have 

 been available in Lower's time, and the choice gener- 

 ally fell on the sheep, partly because of its gentle and 

 amiable disposition and partly " quia Christus est 

 agnus Dei," as Coga said, an indigent bachelor of 

 divinity who subjected himself to the experiment in 

 1667. But transfusion of blood never became an im- 

 portant or popular therapeutic procedure on these 

 terms ; large quantities of foreign blood were found 

 to cause serious and even fatal ill-effects and small 

 amounts did no good. With the discovery of the last 

 thirty years that the tissues of any one species of 

 animal are foreign and more or less poisonous to the 

 economy of any other species came the recognition 

 that transfusion in man could be done only with human 

 blood, and in recent years the value of the procedure 

 has been fully established, large quantities being 

 transfused from a healthy to a sick person without 

 untoward effect. 



In this revival of human transfusion it was, how- 

 ever, soon found that the capacity of the body to 

 identify any blood as foreign to and incompatible with 

 its organisation was based on finer distinctions than 

 zoological species. If from a dozen people a few cubic 

 centimetres of blood are withdrawn, and in each case 

 preparations made of the serum and of the red cor- 

 puscles washed free from serum, and if a sample of each 

 lot of corpuscles is then mixed with a little of each 

 serum in a series of test-tubes, it will be found that the 

 results are not all the same. In some the corpuscles 

 behave as if they were suspended in physiological salt 

 solution — remain dispersed from one another and in- 

 tact ; in other cases they run together into larger or 

 smaller clumps and masses and often disintegrate. It 

 is obvious that the occurrence of this agglutination in 

 the circulating blood is very undesirable, as the masses 

 of corpuscles are liable to block important blood- 

 vessels, and there is plenty of experience to show that 

 serious trouble may be caused in this way. It is 

 therefore not every human blood that is suitable for 

 transfusion into a given person. 



By sorting over a large number of people by this 

 test it has been found that they may be classified into 

 four groups by the satisfactory hypothesis of von 

 Dungern and Hirschfeld. On this view there are 

 two agglutinating factors in human blood serum (a 

 and b) and two agglutinable factors (A and B) in 



NO. 2770, VOL. Iio] 



human blood corpuscles : A corpuscles will react only 

 with a serum, b serum only with B corpuscles. A is 

 never found in the same person as a, nor 1! with b ; 

 either combination would be incompatible with life. 

 The blood characteristics of the four groups are : 



Serum. Corpuscles. 



Group I. . . neither A and B 



Group II. . b A 



Group III. a B 



Group IV. . . a and b neither 



It follows that the serum of Group I. will not ag- 

 glutinate anybody's corpuscles, while the corpuscles 

 of Group I. are agglutinated by all other sera except 

 their own. Group IV. is the reverse of this, while the 

 serum of Group II. agglutinates the corpuscles of 

 Groups I. and III., and the serum of Group III. the 

 corpuscles of Groups I. and II. The corpuscles of 

 Group I. can safely be put only into recipients belong- 

 ing to the same group, those of Group II. only into 

 Groups I. and II., those of Group III. only into Groups 



I. and III., those of Group IV. into anybody. It 

 is a curious fact that in actual practice it is only the 

 qualities of the donor's corpuscles and the recipient's 

 serum which need be considered. When, for example, 

 Group IV. blood is transfused, the plasma of it should 

 agglutinate the corpuscles of the recipient if the 

 reaction took place as it does outside the body. This 

 does not appear to happen, or if it does it produces no 

 obvious ill-effects — which is fortunate, as otherwise 

 safe transfusion would be impossible except between 

 members of the same group. Why this should be so 

 is at present doubtful. It is most probably due to 

 the. quantity of transfused plasma being insufficient, 

 when diluted with the recipient's blood, to cause a 

 significant agglutination of the recipient's corpuscles. 

 The fact that it is plasma which is injected and not 

 serum may also have some influence, though the 

 recipient's plasma has the same effect as his serum, at 

 any rate qualitatively. 



While it is convenient to recognise four varieties of 

 individuals, it will be seen that there are only two 

 factors concerned. A is characteristic of Group II., 

 and B of Group III. ; A + B are present in Group I., 

 and both are absent in Group IV. A corpuscles are 

 necessarily associated with not-a serum, and B corpuscles 

 with not-6 serum. In inheritance these qualities have 

 been shown to be transmitted as straightforward 

 Mendelian factors. It follows that the blood of 

 parents and children are by no means necessarily 

 compatible : though parents both of Group IV. can 

 produce children only of the same group, two Group I. 

 parents may have offspring belonging to any group, 

 according to the particular composition of their 

 hybridity. The possibility of using these blood re- 

 actions to investigate cases of disputed parentage has 

 been carefully worked out by Ottenberg, who shows 

 that the method can have but a limited application, 

 though the answers are conclusive if they can be ob- 

 tained at all. Of much interest also is the observation 

 that the proportion of the population falling into ( Groups 



II. and III. varies a good deal in different races. In 

 England about 40 per cent, are Group II., about 15 per 

 cent. Group III., Groups I. and IV. giving about 2 



