Interesting as the wider questions may be, we are here more immediately concerned with the distribution of the blood groups amongst our own population. The percentages in which the four groups occur have been estimated by various observers, and, as will be readily understood from the foregoing remarks, the numbers show some variation. The approximate figures as worked out by three observers in America are as follows: ------+----------+---------------+---------------- | Bernheim | Moss | Culpepper | | (1,600 tests) | (5,000 tests) ------+----------+---------------+---------------- I | 2 | 10 | 3 per cent. II | 40 | 40 | 38 ” ” III | 15 | 7 | 18 ” ” IV | 43 | 43 | 41 ” ” ------+----------+---------------+---------------- The percentages found among the first hundred men whom I tested in the British Army in 1917 conformed almost exactly to the first of these series of figures, and they may be taken as an average result for Western peoples. It will now be seen upon what grounds it was stated in the last chapter that the chances were in favour of the blood of a donor chosen at random being compatible with that of the recipient. If the patient belong to Group II, then 83 per cent. of other bloods will be compatible. If he belong to Group III, 58 per cent. will be compatible. Only if he belong to Group IV will the chance in favour of compatibility fall below 50 per cent. This statement of the facts concerning distribution of the blood groups will serve to emphasize the absolute necessity for the careful testing of a donor before his blood is used for transfusion. But, further than this, it is necessary to clear away several widely spread misapprehensions as to the group relations between an infant and its mother and between the various members of a family. It has several times been stated in print that a mother’s blood must be compatible with that of her child, or sometimes that a baby has no blood group, so that it may be safely transfused with blood taken from its mother or its father without preliminary testing. On other occasions the statement has been made that the brother or sister of a patient is more likely than other people to belong to the same or a compatible blood group, so that untested blood may be transfused from one member of a family to another with little risk. Knowledge of the existence of blood groups has become somehow mixed up with vague popular beliefs concerning “affinities” and “blood relations.” Such confusions must, however, be dissipated, for none of these statements are more than partially true, and they may lead to a false sense of security and to disaster.