An Exchange in Locard’s Own Words (Part 4)

Translated by Kathleen Brahney
Commissioned by McCrone Associates, Inc.

Edmond Locard
Doctor of Medicine, Professor of Law, Director of the Lyon Laboratory of Police Techniques,
Vice-President of the International Academy of Criminology

Manual of Police Techniques
Third Edition, Completely Revised and Augmented.
Paris: Payot, 16 Boulevard St. Germain, 1939

Part IV—Blood Stains

Chapter IV

The stains one might have to study in the course of a criminal inquiry are of two types—organic and inorganic.  Among the former, we will consider blood, semen, meconium, feces, urine, saliva, diverse secretions and food stains.  Among the latter are stains from candle wax, dyes and mud.



stains are found either at the scene of the crime or on the suspect himself. At the crime scene, care should be taken to search for blood stains on dark fabrics, where they will be difficult to see, particularly on black or red velvet.  One should look carefully at the undersides of table edges, on the arms of easy chairs, on the marble tops of consoles or on the edges of commodes that jut out.  Suspect objects should be seized and taken to the laboratory after scale photos are taken to establish the exact placement of the furniture.  Stains on objects that are not transportable should be photographed and then scraped.  The scrapings should be collected in folded paper, like that used for pharmaceutical powders, and labeled carefully in conformance with an invoice, a copy of which will be included in the statement taken from the accused.  If it is absolutely necessary to choose from among numerous and extensive stains—and when none of them are transportable — one must clear away the ones that are obviously not blood by applying Joseph paper (Tr. note:  Joseph paper, named after Joseph Montgolfier, is thin paper used to filter liquids) dampened in cold water, which will dissolve the blood stains.  One can also use Medinger’s method, which is outlined below.

On the suspect, one will examine the suspect’s clothing, which can only be done in the laboratory.  This will normally involve taking the suspect’s clothes, or at least taking some of the suspect’s clothing and underwear to be examined, along with a careful scraping of the fingernails.

The following questions arise:  

  1. Is the stain a blood stain?
  2. Is it human blood?
  3. What part of the body does it come from?
  4. Which individual does the blood come from?


  1. Taylor’s Prints:  In cases where one is in the presence of a number of large stains, it can be advantageous to eliminate those which are certainly not blood stains.  In order to do that, one can use van Deen’s reaction, which basically consists of turning a tincture of guaiacum blue in the presence of turpentine and blood.  To do this, take a piece of white filter paper or blotting paper, which has been tested to assure that it does not contain any oxidants, which would turn the guaiac blue.  Then add a few drops of fresh tincture of guaiac, a drop of ozonized turpentine essence and a drop of pyridine.  If a blue tint appears after a few seconds, the presence of blood is probable. After that, it is a simple process of elimination, since the blue tint will also be produced in the presence of milk, pus, semen, sweat, starch and numerous plant substances and a whole series of chemical compounds.  In particular, guaiac is turned blue by traces of rust, which makes Taylor’s prints completely useless in the testing of metal for blood.  Moreover, old blood that is heated or spoiled will not turn blue.  This method, therefore, is not perfect, even as a means of elimination or triage.
  2.  Boiling in oxygenated water (Schoenbein):  If you place a fiber of bloodstainedmaterial or a small amount of a scraping from a blood stain on a microscope
    slide and then place a drop of oxygenated water [hydrogen peroxide] on the edge of the slide, you will see a foam of oxygenated bubbles appear.  This method, which is quite simple, has no pretensions to being a positive test because it occurs in the presence of a large number of organic secretions (saliva, pus, semen, sweat, etc.).  It is not even a good method of elimination because the foaming does not take place when the blood is accompanied by the presence of certain salts or acids, or when the blood has been heated.  Nevertheless, since the reaction does occur with old or spoiled blood, this method of elimination is preferable to Taylor’s prints.  In reality, it is reasonable to abandon completely the methods of elimination that are deceptive and use, instead, those methods that yield certitude.
  3. Teichmann’s Method:In distilled water, soak a thread of bloodstained fabric or a small scraping from a bloodstain.  Then, evaporate the liquid over an alcohol lamp or, better yet, on an Ogier heater, being careful to not let the heat rise above 60 degrees.  Once the liquid is evaporated, cover the preparation with a cover glass under which, through capillary action, you will add a drop of acetic acid or, better still, acetic anhydride ((CH3CO)2O).  You will then put the slide back on the heater and evaporate the acid, ensuring that the mixture does not boil out onto the slide.  Above all, make sure that the substance does not dry out completely, and as soon as the liquid is evaporated, add more of the liquid once, and then again; the second time do not heat it. It seems quite useless to add sodium chloride to the primary solution, as Ogier wanted to do, as this addition leads to the formation of solid crystals that mask the reaction.
    The technique just described yields—if the stain is a bloodstain—crystals of hemin or hematin chloride.  The crystals will be elongated, dark or light brown in color, and are often found in twin pairs, with dimensions ranging from 1 to 20 microns. 

    Teichmann’s method lends confusion only in extremely rare cases, such as when there are crystals of murexine present, which are similar to hemin, but which turn from red to violet in the presence of potassium.  Another instance would be in the case of clothing dyed with indigo, which would yield crystals identical to those of hemin. 

    The chances of this happening are negligible, but the flaw in Teichmann’s method is that it is difficult to carry out, and in most cases is so delicate that, even in the hands of a specialist, a negative result must be controlled by carrying out several additional trials.

  4.  Strzyzowski’s method:Prepare a reactive as follows:
    1 cc. each of alcohol, water and ice cold acetic acid, and 3 to 5 drops of hydriodic acid (HI), with a specific weight of 1.5. Then, place a small amount of the dried suspect substance on a slide.  Place a cover glass directly on it.  Using capillarity, add a few drops of the above mixture.  Bring it to a boil for 10 – 12 seconds, replacing the liquid as it evaporates.  In this wayin the presence of a blood stainone will obtain crystals of hematine HI in the form of small, black, rhombic prisms.  De Rechter obtained very clear results even with pickling liquid. This method, which is both quick and very certain, is infinitely preferable to Teichmann’s method.  The disadvantage is that it is impossible to save the reactive agent.  You have to prepare it each time and begin by carrying out a control on fresh blood. De Rechter prepares 6% solutions of HI enclosed in 50 centigram brown glass vials, wrapped in black paper.  The acid keeps for several months.  One doesn’t break open the vial until one is ready to use it.

  5. The Sodium bromide method:The difficulty of conserving HI prompted me to substitute, in its place, hydrobromic acid.  So here is the method used in the Lyon Laboratory of Police Techniques.  Place a very small amount of the dried suspect substance on a slide.  Put a cover directly over it.  Using capillarity, add a few drops of sodium bromide solution of one part to 500. Heat it, but not to the boiling point.  When the evaporation has finished, add, always through capillarity, the following mixture:  1 cc. each of alcohol, water and ice cold acetic acid.Evaporate it again, very slowly.  It is not necessary to get it completely dried.  You will obtain very dark, mahogany-colored, rhombic crystals, which consist of hematine HBr.  These crystals will be voluminous and much larger than those produced using Strzyzowski’s method, and certainly much bigger than Teichmann’s.When the stain is pale and spread out on cloth or bedsheets, instead of removing some threads or taking a scraping and mounting it on a slide, it is better to steep it for a long time in cold water, and concentrate the liquid afterwards—one can then carry out the experiment as described above.
  6. Pierre Medinger’s Method:Pierre Medinger of Luxembourg conceived of a method that, to his mind, would be a simple process of elimination.  He prepares the following reactive:1 gram of leuco-malachite green100 cc. of ice-cold acetic acid

    150 cc. of distilled water

    The reactive can be conserved indefinitely in a flask with an emery stopper and kept out of sunlight.  The chemical name of leuco-malachite green is

    Its chemical formula is:

            C6H4N (CH3)2


    CH – C6H4N (CH3)2


    It is the result of the action of two molecules of dimethylaniline on one molecule of benzaldehyde.  Before using the reactive, one should carry out a test on a old blood stain as follows.  Mix 8 cc of reactive with 2 cc of 1% oxygenated water (or 10 drops of 3% H2O2)   The goal of the trial is to control the oxygenated water, which one is never sure about.  One then places a bit of an old blood stain on a piece of filter paper; one then applies the reactive.  In less than ten seconds one will obtain a green stain which, in less than a minute, will turn a dark bluish-green.

    Once the control is carried out, one proceeds as follows.  If, by using a magnifying glass, one finds a suspect stain on clothing or on an object, one gets as close as possible with a piece of clean filter paper and, using a needle, scrapes off a few grains.  With a glass rod, one then places a drop of reactive on the paper next to the grains.  One should obtain a dark greenish-blue shade, as described above.

    For extremely small stains, one should use a glass rod stretched out as thin as a wire.  With the glass wire, touch the stain and then wipe it on filter paper. One would then add the reactive.

    The control experiments carried out by Jacques Locard at the Police Laboratory of Lyon are of such probative value that one can consider Medinger’s reaction as one of the best and most reliable of existing reactions.  In fact, it comes out negative with all organic liquids except blood, and negative with all the chemical compounds that were examined.

  7. Research on droplets:If the stain is quite recent, microscopic examination will reveal the histological elements of the blood, in particular those of the red corpuscles and their characteristic aspects. If the stain is already a bit old, you can try to regenerate the droplet, either by Virchow’s liquid, which is a 30% solution of caustic potassium, or by using Vibert’s liquid, which contains 2 grams of sodium chloride and 0.5 grams of dichloride of mercury for every 100 grams of water.  Little by little, the red corpuscles will lose their crenellations and take on their primitive form.  The noting of red blood corpuscles dispenses with the need for any further research.
  8. TheFlorence-Nachet apparatus:  The is an apparatus that one can use to adapt an ordinary microscope and which allows for the direct examination of red corpuscles on a stained object.  This method is perfect for searching for blood mixed with rust on metal weapons.
  9. Spectroscopic method:  In principle, the use of the large hemato-spectroscope is reserved for cases in which the abundance of stains allow for the use of sufficiently concentrated solutions.  On the contrary, one can always use the spectromicroscope no matter how small the stain.  The goal is to determine the presence of bands of hemochromogen or hematin reduced in an alkaline solution. 

    One takes a speck of blood — or one might dissect the cloth — using a platinum needle (not iron, since rust would prevent the reaction from taking place); place the material onto a slide and place the focus point over the darkest spot.  By capillarity, introduce a drop of ammonium bisulfide.  The focus should be set carefully, using a strong lens; one then removes the eyepiece and replaces it with a spectroscope.  If the spot is a blood stain, one will see two bands of from hemochromogen appear.  The first will be very dark and quite clear between D and E, the second will be thin and will appear a little beyond E. 

    Apart from stains mixed with rust, using the spectroscope is perhaps the best choice for determining blood.


The question of whether a blood stain is of human origin or from some other animal species has been treated using a variety of methods.  The examination of red blood cells and measuring them has long been the only known procedure.  Since 1895, one has come up with a series of new methods — precipitating serums, erythroprecipitation and erythroagglutination.  Among these, I will mention only the deviation from the complement and the anaphylactic methods as being the most reliable (Dervieux and Leclercq.  Diagnosing stains in Legal Medicine.  Paris:  Bailliere, 1912.)

  1. The examination and measuring of blood corpuscles.  The red corpuscles are elliptical in birds, reptiles, batrachians [tailless amphibians] and fish, and also in camelids (camels, dromedaries, llamas, alpacas and vicunas.)   Thus there is an immediate and simple process of elimination when the stains, for example, are left by feathered game or chickens.  All mammals have round corpuscles, but the diameters vary.  One can make a diagnosis, therefore, by reconstituting the corpuscles as described above and by making various measurements and determining the mean.The findings:


  .0069 – .0077 mm


  .0066 – .0073 mm


  .0060 – .0069 mm


  .0060 – .0065 mm


  .0058 – .0065 mm


  .0056 – .0060 mm


  .0055 – .0057 mm


  .0040 – .0046 mm


  .0040 – .0046 mm

If the corpuscles are deformed, the results are very uncertain.  And determining monkey blood is impossible.

  • Deviation from the complement (Bordet and Gengou):The principle is the following:  When one injects red blood corpuscles of goose blood into the
    peritoneum of a guinea pig, those corpuscles are devoured by the phagocytes. The blood of the guinea pig thus takes on the ability to hemolyze the red blood cell of the goose in vitro, an action which the blood was incapable of doing previously.  The hemolyzing property thus acquired is due to the simultaneous formation of two substances in the blood of the sensitized animal.  One is called the sensitizer or amboceptor, which only acts on the red blood cells of the species to which it has become sensitized (in this particular case, the goose.)  The other is called alexine or the complement, which is the same no matter what the species is that the corpuscles have acted on.  It is the sensitizer that allows the alexine to fix itself to an antigen, that is, to the species possessing the foreign blood corpuscles.  Stated another way, under the influence of the sensitizer, the antigen acquires the capacity to fix alexine, or—as Bordet and Gengou put it—to become the complement.


That said, here is how the reaction takes place.  “Let’s suppose that A is an antigen [e.g. human blood] and that B is the corresponding sensitizer [for human blood].  If one puts a sufficient quantity of alexine in the presence of A and B, the alexine, thanks to the intermediary B, will fix to A.  There will be no free alexine left over.  Then, if one adds goat corpuscles and a hemolytic serum for these corpuscles [test reagent cells and serum], heated to 56 degrees to make the alexine disperse, the goat corpuscles, while preserving their sensitivity, cannot be hemolized, for lack of free alexine, and Bordet and Gengou’s reaction will be positive [for human blood].  On the contrary, suppose that A does not correspond to the sensitizer B.  The alexine added will not fix onto A.  It will remain free and will be able to fix on the goat corpuscles thanks to the corresponding sensitivity of those corpuscles.  Hemolysis will take place and Bordet and Gengou’s reaction will thus be negative [for human blood].  (Dervieux and Leclercq)

This method is excellent, except in the case of monkey blood, where it lends confusion. But the process is extremely difficult and long.  Technical errors can destroy the value of the result.  It necessitates a laboratory equipped with special tools.  This test can only be carried out by a practiced professional.

  • Anaphylaxis (Uhlenhuth, Minet and Leclercq):Anaphylaxis, which is the opposite of immunity, is the particular state of susceptibility of an animal to a dose of a substance which it has been injected with previously.  Thus, if one sensitizes a guinea pig by injecting it with human blood, it will be overwhelmed by a new injection made 20 days later with the product of the dilution of a stain, if the stain consists of human blood.


In practice, one dissolves the stain in a very small quantity of physiological salt water to which one has added a weak solution of caustic soda in order to render the solution slightly alkaline.  Heat it to 100 degrees in a bain-marie to rid it of microbes.  One then sensitizes a series of guinea pigs through intra-cardiac injections of 1 cc of the solution.

Fifteen or 20 days later, one gives one guinea pig an intra-cardiac injection of 1 cc of human blood.  If the stain is of human blood, the guinea pig will be overcome. If the result is negative, one can then inject the other guinea pigs with blood from dogs, rabbits, horses, etc. until one obtains the anaphylactic reaction to determine the origin of the blood stain.This method is much simpler than the deviation from the complement and much more reliable. It is only inapplicable in the case of monkey blood.  The only prerequisites are a conventional installation and sufficient training.


Dervieux proposed a method for determining the individual origin of blood.  Every three days, one gives a rabbit a subcutaneous injection of 2 cc of pure, fresh human semen, with live spermatozoa.  One injects a total of five doses.  At the end of three weeks, one bleeds the animal from the carotid artery, collects the serum and preserves it in vials.

This serum will yield precipitations with human semen and with human blood, whereas the serum prepared with blood only reacts with blood.  It precipitates with very weak solutions of human blood, whereas serum made with blood gives no yield. It precipitates with blood from men but does not precipitate even with more concentrated dilutions of blood from women.  It gives a more intense precipitation with the blood of the same individual than with the blood of any other individual.

One can see that diagnosing the individual rests on the difference of intensity of the reaction.  The interpretation is thus not easy, and there is room for reservation in drawing one’s conclusions.

Since the creation of this test, the question of individual origin has made appreciable progress with the discovery of “blood families.”  In the human species, there are four different kinds of individuals according to blood type.  With the aid of this diagnostic tool, one has acquired a new means to resolve the problem of paternity.  But in terms of identifying stains, one must be very suspicious of methods that are only useful in the case of fresh and abundant blood.

On this point, it is useful to consult the work of Leon Lattes, notably his book The Individuality of Blood in Biology, Clinics and Legal Medecine.  Paris, Masson, 1929.   Here is an essential passage borrowed from a recent work of the eminent specialist, Leon Lattes:

“Landsteiner succeeded in establishing that human blood has a strictly individual biochemical physiognomy.  Certain of these individual characteristics have been placed in evidence by the discovery of agglutinogenous properties revealed by immunizations and procedures independent of the classic groups.  With ease, Landsteiner was able to identify the blood of ten or so individuals working in his laboratory.  From an experimental point of view, the problem of strict individualization can be considered resolved.  From a practical point of view, it is quite the contrary in the case of blood stains, since it is not certain that in the case of stains, one can regularly recover all the signifying properties for which—in general—studies have scarcely begun.  It would be, without a doubt, very dangerous to apply the results of experiments to criminology which, while extremely interesting, may not yet be reliable in yielding absolutely constant results.”

Here, however, is the technique to follow in determining blood types.  It is quite simple, provided that one can procure serums II, III and IV.  One has had excellent results in [the crime labs of] Lyon, Bern and Berlin, among other cities.

There are four blood types, designated as follows:

I or AB, II or A, III or B, IV or Zero(O)

The particularities of the groups depend on the properties of their serums and above all on the properties of the red blood cells.

The simplest and most reliable method of determination is as follows:  Using a syringe, one takes a blood sample from the crook of the arm or from a finger of the subject.  Separately, one places on a glass slides, two drops of serum each from groups II, III and IV.  Using a glass rod, one mixes a drop of blood from the subject with drop of serum, and looks at it immediately.  If there is agglutination, it will appear in 2 – 3 minutes in the form of a mass of red cells.

The reaction is quite clear; it cannot be confused with sedimentation—for which one has the deposition of regular piles and then, later, coagulation followed very quickly by a retraction of the clot.  In agglutination, truly irregular masses form, appearing well before coagulation occurs.  It is good to carry out this experiment in an ambient temperature of about 16 degrees C.

The following eventualities can occur:

  1. Agglutination with all the serums:  the individual is type I
  2. Agglutination with serums II and IV:  the individual is type III
  3. Agglutination with serums III and IV:  The individual is type II
  4. No agglutination:  The individual is type IV


The suspect who has blood on him often says that he has had a bloody nose or that the blood is from an accident.  It is important to verify these statements.  First, one notes whether the shape and the placement of the stains contradicts the statements.  A blood stain on the back of an apron cannot be from a nosebleed (epistaxis).  Multiple small, round drops are the result of projection.  Blood that stains the outer side of underwear more than the interior have a very little chance of being catamanial [sic.]  In contrast, tear-shaped stains on a front of a vest or jacket could very well be from a nosebleed.

Sometimes, but rarely, microscopic examination can yield information about the origin of a bloodstain.  In blood from a nosebleed, one can find hairs and epithelial cells from nose hairs.  In hemorrhages resulting from rape, one may find semen and pubic hair.  In menstrual blood, one may find cells of uterine mucous and certain specialized  microbes, such as vaginal trichonoma.  In blood that is spit up, one finds food remains and cells from the digestive tract.  In blood stains from flea bites, one finds black granules excreted by the insect, etc. It is well understood that, if the presence of such elements leads to a clear affirmation, their absence in itself is not conclusive.

If it is important to state the quantity of the blood that has been shed, one must guard against the  common tendency, which always is to overestimate the quantity. One can weigh the crusts or flakes of blood.  For stained objects, one can determine the weight of the blood stains by noting the difference between the weight of the object before and after it has been washed, calculating that the dry stain weighs one-fifth of the weight of fresh blood.  For blood found dissolved in wash water, one should keep in mind that a few drops of blood are sufficient to turn a liter of water red.


It is sometimes very important to determine the precise date that a bloodstain has been made.  Sometimes the circumstances surrounding the stain can be helpful—the presence of dust on the stain, the fact that the stains are on freshly cut grass or on dead leaves.

Also, some substances hasten coagulation (lime salts, foreign albumin, peptones, iron perchloride, tannin, lemon juice, hot water); others slow it down (sodium citrate, benzene, arsenobenzol.)  Heat activates the process; cold slows it down.

A stain that contains red cells is not very old.  If the red cells are not deformed, the stain is recent.  If one can see in the spectrum a third band between C and D (methemoglobin), the stain is old.  The speed of the dissolution of the albumins is inversely proportional to the age of the stain.  (Mutrux, “Determination of the Date and Quantity of Blood Stains,” in the Internatinoal Review of Criminalistics, 1938, no. 2.)


One always has a tendency to overestimate the quantity of blood in stains.  It is true that dried blood loses 80% of its weight.  Therefore, the  weight of blood on a non-porous surface is five times that of the stain.  The approximate volume of the blood is equal to the weight of the stain divided by 1.04.


add comment