JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
Back to the library
Medicine 1901

On Agglutination Phenomena of Normal Human Blood

Karl Landsteiner

Healthy people's blood comes in distinct, incompatible kinds — and that is why transfusions sometimes killed.

Choose your version
In depth · the introduction

For centuries, giving one person another's blood was a gamble that sometimes saved a life and sometimes ended one — until a young doctor in Vienna found that human blood comes in distinct kinds that don't mix.

The big idea

Karl Landsteiner noticed something nobody had taken seriously: if you mix the clear liquid (serum) of one healthy person's blood with the red cells of another, the cells sometimes clump together — and sometimes don't. It depends on who is matched with whom.

Testing the blood of 22 healthy people against each other, he saw a clean pattern: people fall into a few groups. He called them A, B and C (we now call C the O group). Your group decides whose blood your body will accept and whose it will attack. That single insight is why a blood transfusion today is routine instead of a coin-flip with death.

How it came about

By 1900, transfusion had a grim reputation. Sometimes the patient recovered; sometimes their blood clumped and they died, and no one knew why. Many doctors assumed the clumping was a sign of disease. Landsteiner, working in a Vienna pathology institute, suspected something stranger — that healthy people simply differ from one another.

So he did the obvious-in-hindsight experiment: he and his colleagues drew their own blood and crossed every sample against every other. The clumping wasn't random. It sorted everyone into a small number of types, with a tidy rule — and your own blood never clumped against itself. He published it in 1901 in a three-page paper. The fourth and rarest type, AB, was spotted by two of his colleagues the very next year.

Why it mattered

This is the discovery that made blood transfusion safe. Once you can type a donor and a patient, you can match them — and the operating theatre, the maternity ward, and the battlefield hospital all became survivable in ways they had never been. Landsteiner also pointed, correctly, to two other uses: identifying blood at a crime scene, and finally explaining why past transfusions had been so unpredictable. He won the Nobel Prize for it in 1930.

A way to picture it

Think of each red cell as wearing a coloured flag: A cells fly an “A” flag, B cells a “B” flag, AB cells fly both, and O cells fly none. Your immune system is a bouncer who attacks any flag it doesn't recognise from home. An O person has seen no flags, so their guard attacks both A and B — but their own flag-less cells offend no one, which is why O blood can go to anyone. An AB person has seen both flags, so their guard waves everyone through. The flags are the antigens; the bouncer's grudges are the antibodies in your plasma.

An interactive blood-typing card: choose group O, A, B or AB and the same red cells are dropped into two wells — one of anti-A serum, one of anti-B. The cells clump in a well only when they carry the matching antigen, so the clumping pattern reads back the blood group.

Where it sits

Landsteiner stands at the start of immunology's practical age. Behind him are Pasteur and Koch, who showed disease could be tamed; ahead of him are the discovery of the Rh factor (which he also helped find, in 1940), modern blood banking, and organ transplantation — every one of which depends on matching donor to recipient. In the Library you can read Fleming on the next great life-saver, the antibiotic; together they made twentieth-century medicine survivable.

The original document
Original source text

Opening — blood differs between healthy people

Karl Landsteiner · Wiener klinische Wochenschrift 14 (1901): 1132–1134 · English translation
Some time ago I observed and reported that blood serum of normal human beings is often capable of agglutinating red blood corpuscles of other healthy individuals.
A real difference between my observations and Maragliano's is that in Maragliano's case the serum acts also on corpuscles which come from the same individual and that the reaction is obtained only with abnormal blood. My observations, however, show differences of a rather striking kind between blood serum and corpuscles of different apparently entirely healthy human beings.

The agglutination tables

Tables I–III · sera of apparently healthy persons crossed against one another's corpuscles
About equal amounts of serum and approximately 5% blood suspension were mixed in 0.6% saline solution and observed in hanging drops or in test tubes (the plus sign denotes agglutination).
All 22 examined sera from healthy persons gave the reaction. The result obviously would have been different had I not used a number of different corpuscles for the test.
[ … ]

The division into three groups

Landsteiner sorts the 22 sera into three regular classes — the groups later named A, B and O
In addition, however, a remarkable regularity appeared in the behavior of the 22 blood specimens examined. … in most cases the sera could be divided into three groups:
In several cases (group A) the serum reacted on the corpuscles of another group (B), but not on those of group A, whereas the A corpuscles are again influenced in the same manner by serum B. In the third group (C) the serum agglutinates the corpuscles of A and B, while the C corpuscles are not affected by sera of A and B.
In ordinary speech, it can be said that in these cases at least two different kinds of agglutinins are present: some in A, others in B and both together in C.

The closing lines — forensics and transfusion

The described agglutination can also be produced with serum which has been dried and then dissolved. … Thus the reaction may possibly be suitable for forensic purposes of identification in some cases, or, better, for the detection of the nonidentity of blood specimens…
Finally, it must be mentioned that the reported observations allow us to explain the variable results in therapeutic transfusions of human blood.
Karl Landsteiner · Vienna · 1901