The method of securing the blood is the same for all the newer methods of determining hemoglobin.

    It is important to secure the drop of blood in such a manner as to prevent its changing from its condition within the vessels as little as possible.  The lobe of the ear, the palmar surfaces of the fingers and the balls of the toes are almost devoid of sensory nerves.  The skin at the sides of the fingers is thinner than that upon the heavier parts of the balls and is often less sensitive.  In persons of certain occupations, such as piano or violin playing the finger tips may have thick skin.  It is necessary to avoid any area which is more than usually liable to injury after the prick is made, the violinist might find some difficulty in playing with the injured finger for a few hours after the prick has been made, for example.  The lobe of an ear is the best site in adults, the side of a toe is best in babies.

    Having selected the site of the puncture, the skin should be cleaned in any way that is convenient.  Probably a good washing with clean water is best for ordinary conditions.  Various antiseptics are sometimes employed, but these must be sufficiently dilute to prevent irritation to the skin; this means little or no antiseptic value.  Any of the ordinary antiseptics strong enough to injure bacteria injure t he tissues still more seriously.  But a washing with anything which removes the dust and perhaps a few desquamating epithelial cells takes away the chief source of infection and does not prevent the most speedy repair.  Care must be taken in the washing that the skin is not irritated enough to cause overfilling of the vessels since this modifies the hemoglobin content.

    The first drop of blood must be wiped away, and the second or later drops taken up in the manner suitable for each instrument.  The tissues must not be squeezed or handled roughly but the blood musts come gently and freely from the puncture.

    The puncture is best made with a sharp, angular lancet.  The lancet which is provided with a spring makes a sudden thrust which is not painful at all.  A pen which has been broken longitudinally in half, so that only one sharp point remains, is useful.  A dozen or more pens can be prepared and sterilized and a new one used for each patient.  Glove needles, being triangular, make a sharp, clean puncture which heals at once.  Round needles are not quite so suitable because they cause more pain and leave a rougher puncture; the slight bruising which occurs may affect the concentration of the blood in the vicinity of the cut.  The error due to this factor is negligible in most cases.

    The blood drop as it wells up from the tiny wound made for the purpose gives a general idea of its  hemoglobin content, by its color.  When the blood is more viscid, this drop stands up round and high, and appears to be richer in hemoglobin than it really is, when it is of low viscidity it spreads out over the surrounding skin in a thinner layer, and appears to be lower in hemoglobin content than it is.  The character of the skin modifies the height of the drop, also, as does the size of the wound and the rapidity of the flow.

    The simplest, cheapest, and easiest method of estimating the hemoglobin is by means of the Tallquist Hemoglobin Scale.  It consists of a scale of ten colored paper slips for comparison, each corresponding to a certain percentage of normal hemoglobin, and strips of filter paper of uniform thickness to take the blood.  In using this scale the drop is secured by the usual means, and a strip of the filter paper laid gently and quickly over the drop, until the blood is soaked up into the paper over an area about one-fourth of an inch in diameter.  This is held for a few seconds or until the first shiny look of the blood stain has disappeared, then it is compared with the color scale.  Some of the scales have an opening in the center of each of the ten color-tint blocks, and the blood-stained paper is then placed beneath these openings, one after another, up and down the scale in succession until the matching tint of the scale is found.  If the blood is apparently half way between two blocks as 70% and 80%, the hemoglobin may be considered as 75%.  In this way, those who are experienced in the test can make estimations of the hemoglobin which are fairly useful though they cannot be considered accurate within more than 20% under the best of circumstances in the most skillful hands.  The method is perhaps the most convenient of all, since there is no apparatus to wash, no special light requirements are necessary and the scale is so small that it can easily be carried in the pocket.  Its lack of accuracy is the only factor against its use.  The method is useful for a preliminary test or in emergency cases.

    Next in ease of use is Dave’s Hemoglobinometer.  This instrument has a revolving color scale of glass, a telescoping tube for securing distinct vision, and two slips of glass which hold a measured film of blood between them.  The slips of glass are placed in the holder and screwed closely but not too tightly together; the blood flows between them by capillary attraction.  The holder than slips into its place at the back of the instrument.  The telescope is pulled out until the edges of the two circles show very plainly and in good focus.  The revolving color scale is moved by means of the screw at the upper part of the instrument, until the tint of the two circles appears exactly the same.  The amount of hemoglobin is read off in percentages of the normal from a scale which is attached to the side of the glass circle, and which is seen through a tiny window at the side of the instrument.

    In using the Dave’s instrument it is necessary to have a yellowish light, not too brilliant, and to make the examination in a rather dark room.  It is better to look with the eyes alternately, thus giving each eye an opportunity to rest, in turn.  The correctness of the reading may be gauged by the fact that one is able to return to about the same figures several times in succession, after changing the position of the scale and resting the eyes.  It is probably correct within 5% of the actual amount.  This scale is made so that blood containing 137.7 grams per liter is considered as 100%, or normal.  It is evident that in using this instrument correction must be made for the age of the patient.

    Gower’s hemoglobinometer is not now in general use.  It consists of three tubes, one filled with gelatine stained with picrocarmine to be used for artificial light, one filled with gelatine stained slightly differently, to be used with daylight, and one to be filled with diluted blood, the latter graduated.  The graduated tube is filled to a certain point with blood, then water is added until the tint exactly matches the tube filled with gelatine, using the one or the other according as the light is artificial (yellowish) or is daylight.  The dilution is difficult; the gelatine tubes fade quickly if they are exposed to the light often (as they must be if they are used at all frequently) and the newer instruments are much more accurate and convenient.

    Sahli’s hemoglobinometer is somewhat like Gower’s but has certain advantages over the older instrument.  It consists essentially of two tubes, one filled with a stock solution of 1% acid hematin, and one tube for the blood.  Twenty cubic millimeters of blood are taken into the blood tube, and this is mixed with deci-normal hydrochloric acid.  (15 c.c. HCI:  1 liter, H2O is sufficiently accurate.)  The acid changes the hemoglobin of the blood into the stable acid hematin.  The blood mixture is then to be diluted with water until it matches accurately the color of the stock solution in the other tube  The advantages of this instrument are that it can be used in any light; since both the stock solution and the blood mixture are colored by the same compound—acid hematin,--they are equally affected by different lightings, so that yellow light, daylight, dim lights or brilliant lights all act upon both tubes in the same way.  A moderate light, however, gives the most nearly accurate readings.  This instrument is, in the hands of reasonably skillful observers, accurate within about five per cent.  The readings are made in percentages of the normal, which is thus subject to the necessity for correction for the age of the patient.  In this instrument 172 grams of hemoglobin per liter of blood are taken as 100%.

    The most accurate, the least convenient and the most expensive hemoglobinometer is that of Fleischl as modified by Miescher.  This has for its scale a long, slender, delicate glass bar which is thick at one end and thin at the other, so colored as to be spectroscopically identical in color with hemoglobin.  This has a scale connect ed with it, and is set beneath a stage which is perforated in its center.  A second perforation near the edge of the stage permits the scale to be read easily.  The colored prism is moved across the central perforation.  Beneath the prism is a reflecting dull white surface and a candle or other dim yellowish light is placed in front of this dull reflector.

    The blood is taken in a pipette somewhat like those used for blood dilution for counting, except that the hemoglobinometer pipette is so graduated that dilutions of 1-200, 1-300, or 1-400 can be made.  Dilution is made with 0.1% sodium carbonate; ordinary tap water is really better, in our experience.

    If the blood appears very pale as it emerges through the puncture use the dilution 1-200.  If it appears almost or quite normal in tint use a dilution of 1-400.  For milder anemias uses 1-300 dilution.  In taking the blood into the pipette keep the lower end of the pipette within the blood drop but do not allow it to touch the skin.  If by accident the blood should be drawn above the selected mark on the tube, or should not quite reach that mark, there are accessory graduations by means of which adequate correction can be made after the work is completed.  Draw the diluting fluid into the pipette to the mark at the top of the bulb, and then immediately thoroughly shake the pipette from side to side.  The shaking must be gentle, and the pipette must not be shaken endways lest some of the blood be forced into the capillary portions of the pipette.

    Two chambers, each divided into compartments, are used for the examination.  Each chamber has a glass cover and a glass bottom, and a perforated dark metal cap to be placed over the cover glass.  One chamber is 12 millimeters in depth, the other 15 millimeters, and readings are made with both chambers.

    Put the glass bottom into the 15 millimeter chamber and fix it in place by means of the screw-like arrangement provided for the purpose.  Fill one compartment with water or the solution used for dilution.  By means of a rubber bulb blow out and discard four drops from the pipette used for diluting the blood.  (This removes the fluid in the capillary bore of the tube, which presumably contains no blood.)  Still using the rubber bulb, fill the other compartment with the diluted blood.  Be very careful that both compartments are completely filled, that none of the clear fluid reaches the blood compartment, and that none of the diluted blood reaches the clear fluid.  Cover with the glass; there must not be any air drops in either compartment.  Place the metal cap over the cover in such a manner that two equal squares are visible, one of the diluted blood, the other of the clear fluid.  Place the chamber in the perforation provided for the purpose in the stage of the hemoglobinometer in such a manner that the clear square is exactly over the colored bar.  Arrange the candle and the reflecting surface so that a dim light is thrown upward through the colored prism and the clear fluid, and through the diluted blood.

    Move the prism back and forth by means of the screw until the two squares have exactly the same tint.  Use first one eye then the other.  Do this quickly; if too much time is taken vision becomes less acute.  Read and note the figures on the scale.  Move the prism away from matching tint.  Close the eyes for a few seconds, then repeat the test.  In this way make ten different readings.  If the readings do not vary more than five points the readings are sufficiently accurate.

    Fill the 12 millimeter chamber in exactly the same manner and make ten readings in the same manner.

    Make an average of the ten readings from the 12 millimeter chamber and make an average of the ten readings from the 15 millimeter chamber.

    Divide the first average (from the 12 millimeter chamber) by 4 and multiply by t, this should equal the average made from the ten readings from t he 15 millimeter chamber.  If the figures thus secured do not agree within five points the work is not sufficiently accurate and should be repeated.

    Compare the figures secured, corrected for the 15 millimeter chamber, with a scale provided with the instrument.  This gives the grams of hemoglobin per liter of the diluted blood.  If the blood was diluted 1-200, multiply the scale reading by 200, which gives the grams of  hemoglobin per liter of whole blood.  If the dilution was 1-300 or 1-400, multiply by these figures.

    This instrument requires an absolutely dark room; it is too cumbersome to carry about easily; its use requires much time for the actual work and for the washing of the apparatus.  Parts are easily broken and they cannot be easily replaced.  It is by far the most accurate instrument on the market and for this reason has an important place in hospitals and research laboratories.  We use it for checking up on other simpler instruments, and in cases in which unusual accuracy is required.
 
 

    In using Dare’s hemoglobinometer, the slides may not be thoroughly dry, especially if it has been necessary to clean them after errors in technique.  The slightest  film of moisture upon the surfaces of the glass slides affects the reading by diluting the blood.  Sometimes the slides are not screwed closely together; the film of blood is then too thick, and the reading too high.  Sometimes the slides are screwed too tightly together, whereupon they break.

    The finger should be removed from the wheel which turns the scale, before the telescope-tube is removed form the eyes.  Else, when the telescope is taken from the eyes, the fingers which have been turning the scale may still turn it slightly.  The scale may thus be turned and the reading correspondingly inaccurate.

    The use of Sahli’s or Gower’s instrument depends for its accuracy upon the care with which the blood is taken, is blown into the larger diluting pipette, and with which the readings and dilutings are performed.  For if the drop of blood should not be removed from the end of the tube, after the requisite 20 c.c. have been taken, a part of this is almost sure to be drawn into the tube with the diluting fluid, and the determination is increased.  If the tubes are moist, the blood is unduly diluted, and the readings are too low.  If the blood is not thoroughly rinsed into the mixing pipette, the readings are thus too low.  In adding the last of the water to the blood in the mixing pipette, it is best to read off the hemoglobin percentage, and write it down, before the later and decisive readings are taken.  Then, if by accident or by choice, the dilution should be sufficient to carry the color past that of the control, the reading can be corrected accordingly.

    In using the original Fleischl instrument sources of error are plentiful and for this reason it is not now used in many laboratories.  The Miescher modification is subject to much less error but still it is necessary to use great care in making every step of the procedure.  In very important cases for research work it is our practice to take the blood in two different dilutions and to compare these; discarding the results if the results do not agree within five points of the final computation, and making an average of the two final results if they do so agree.

    For each method of determining hemoglobin accuracy is gained only by experience and carefulness; with these the hemoglobin can be determined reasonably well, within the limits of accuracy for the instrument employed, in any case.
 
 

    The lighting is important.  In our laboratories a small electric light placed squarely beneath the condenser is used for ordinary work.  A larger light placed in front of the small round reflecting mirror is good; this permits variations in lighting for low-power or high-power magnification in addition to the modifications secured by regulating the iris diaphragm.

    Other articles needed are a lancet for pricking the skin, various solutions and containers, a counting chamber, pipettes for diluting the blood and sterilizers and other articles ordinarily present in a laboratory for clinical diagnosis.
 
 

    This ruled area is upon a part of the slide which is elevated above the rest of the slide; the exact amount of the elevation is of no consequence, and varies for different counting chambers.  Around this ruled glass area, which is, properly, the chamber, there is a moat.  This is a depression around the ruled area, itself surrounded by another elevation of glass.  The glass around the moat is always 1-10 millimeter higher than the counting chamber, so that when the cover-glass is placed upon the slide a space exists between the ruled area and the cover-glass; this space is 1-10 millimeter in depth in all chambers in ordinary use.

    The counting chamber requires care in cleaning because the ruled lines are delicate and t he glass can be scratched easily.  If the chamber is made by cementing different pieces of glass together, the chamber must never be washed or rinsed with anything but warm water.  This is really all that is necessary, anyway, in ordinary cases.  If the counting chamber is made of a single piece of glass, cut to form the various parts, it may be washed as any delicate glass-ware would be.

    The ruled area must be treated with especial care, it must never be rubbed in drying, but only very gently mopped, or, better, left to dry after a final rinsing in warm distilled water.
 
 

    The tube frequently used for counting the white cells has a smaller bulb, or a larger bore in its capillary tube, so the bulb contains only ten times as much fluid as the capillary tube.  This tube also is divided into ten equal parts.  The fifth division is marked .5 and the upper division is marked 1 as in the red cell pipette, while the upper limit of the bulb is marked 11.  Thus it is possible to secure dilution of 1-10, 1020 or even 1-100 by drawing the blood to the 1 mark, to the .5 mark, or to the lowest mark, which is 12-10 the content of the bore of the capillary tube. Other markings are employed for the white cell pipettes occasionally; their significance is easily understood by a comparison of the markings with one another and with the red cell pipettes.

    For both pipettes there is a dilated area above the constriction which marks the upper limit of the bulb; this receives any superfluous amount which may be drawn up beyond the proper limit, by accident.  There is a rubber tube with a glass mouth-piece for each blood pipette, and it is best to have these rubber tubes a foot or more in length.  The rubber tubes which come with the pipettes are rarely long enough to permit accuracy in vision or delicacy in manipulation of ascending columns.  By having the longer rubber tubes it is easier to see clearly the ascending column of blood, and the increased elasticity due to the longer rubber tube allows greater accuracy in manipulating the columns as they are drawn upward.
 
 

    If coagulation of blood should occur within the capillary bore or the bulb, this must be digested out.  An artificial gastric juice containing both free hydrochloric acid and pepsin, or an artificial pancreatic juice containing trypsin and mildly alkaline in reaction, should be drawn into the bulb and the pipette left in an incubator over night.  If the clot is not dissolved the process must be repeated, perhaps using some other artificial digestive fluid. After the clot has been digested the pipette is rinsed with cool water, then with hot water, then is dried in the usual manner.

    A filter pump attached to the water tap cleans the tube very easily and efficiently.  Attach the pipette to the intake of the filter pump, place the other end of the pipette in cold or hot water, or in the acid solution if this is to be used.  Turn on the water and allow the cleaning solution or water to flow through the pipette as long as seems desirable.  Then remove the lower end of the pipette from the water and allow air to flow through until the pipette is perfectly dry.

    If the filter pump is not at hand, the pipettes must be cleaned by hand.  Cold water, then hot water, then 95% alcohol, then ether are used in turn and then air is blown through the pipette until the tube is dry.  Draw the fluids into the pipettes by suction through the rubber tube, then expel them by using a rubber bulb.  Air is drawn into the pipettes and then is expelled by the rubber bulb.  The final rinsing with alcohol and ether is intended to hasten the drying process because these fluids evaporate rapidly.  (They are not necessary when the filter pump is used because the time question is not important to the filter pump, and hot water is not usually limited.)  The cleaned pipettes may be kept in any place which is free from dust and moisture.
 
 

    Lancets which are provided with a spring are rather less easily cleaned.  They have the advantages of producing wounds of equal depth, and they can be set to make deeper wounds if the skin shows marked pallor, or shallower wounds if the patient is a bleeder or if the skin is very red.  The needles without a spring depend upon the skill of the operator for the accuracy of gauging the depth of the wound.

    Needles and lancets are best sterilized by rinsing in warm water, then dipping them into carbolic acid.  After another rinsing in sterile water they may be dried on sterile cotton or left to dry.
 
 

    For the red cells many solutions are employed.  In our laboratories normal salt solution is tinged with a few drops of methylene blue and this is used for diluting the blood for the red cell count.  The cells remain in fairly normal condition for several hours, which is all that is required for ordinary cases.  If the blood must be kept in the pipette for some time before it can be counted, Hayem’s solution is excellent, and this is the one most used.  The following is the formula;

    Distilled water, 100 c.c.
    Sodium chloride, 1 gm.
    Mercuric chloride 0.5 gm.
    Sodium sulphate 5. gms.
    Another solution which is preferred by many workers is Toisson’s fluid.  Its formula is:
    Distilled water, 160 c.c.
    Neutral glycerine, 30 c.c.
    Sodium sulphate, 8. gms.
    Sodium chloride, 1. gm.
    Methyl violet, 25 mgs., or just enough to give a faint purplish tinge to the solution.  This fluid stains the white cells.  It does not keep so well s Hayem’s fluid, does not fix the cells appreciably, and we have found it less satisfactory than Hayem’s fluid for keeping the blood for a long time (a day or two or three days)under experimental conditions.
 
 

    For cases requiring efficient sterilization the solutions to be employed are selected according to the nature of the infectious agent suspected.  The methods of surgical procedure are followed, and this sterilization must precede the taking of the blood by an hour or by several hours, according to the manner in which the skin reacts to the sterilization process.

The use of alcohol, ether or any solution which reddens or which pales the skin, or which causes any sensory irritation worthy the name is bad and interferes with the accuracy of the count.

    Cotton or gauze pads for applying the cleansing solutions should be sterile and at least five pads of gauze or bits of cotton should be ready.
 
 

    The side of a toe is the selected site for taking blood from babies or small children, for the following reasons:  Vaso-constrictor control is not abundant; the skin is thin; the foot is easily held, if the child is awake, and the toe is not apt to be irritated or infected afterward.  If the child is asleep the process may not awaken him.

    The lobe of the ear is selected for adults for the following reasons:  The lobe of the ear is not visible and the patient cannot see the blood; the vaso-constrictor control is not abundant; the skin is thin, sensory nerves are scanty, and the lobe of the ear is not subject to later irritation.

    In the case of an adult the worker stands at the left side and to the rear of the patient.  The patient sits comfortably by the side of a small table upon which the necessary equipment is placed conveniently for the worker.  The patient cannot see the blood being taken.  If the patient is in bed he should turn the face away from the worker, thus permitting an ear to be accessible.

    If really efficient sterilization of the skin should be necessary, this should have been done at least an hour before the blood is to be taken, and the skin protected by cotton or gauze in the interim.  In most cases the washing which removes dust and some desquamating epithelium is all that is necessary.  The ear lobe should be gently mopped with sterile water, dried, and protected against dust if any delay should occur.  The lobe must not be rubbed or handled or caused to redden perceptibly by the cleansing.  Such cleansing agents as alcohol are to be avoided because they dilate the capillaries and modify the counts.

    If the patient has come in from the street a mild soap solution may be used for cleaning, then the ear washed with sterile water and dried with sterile cotton.  If the cleansing should cause visible reddening of the skin the ear should be protected with sterile cotton and the patient allowed to rest for ten minutes or more, until the normal circulation has been well established in the ear.

    Prick the skin with lancet or needle.  Notice the emergency of the first drop, its size, color, and the manner in which it flows upward into a round drop or spreads around over the skin.  Wipe away this first drop.  Take the blood for the white cell count first.  This is partly because the white cells stand longer without being modified, partly because the white cells are more quickly modified by the faint vasomotor change due to the slight irritation of the prick, and partly because a larger drop is used for the white cell count.  (The first drops are usually rather larger than later droops of blood from so small a prick.)  If the white blood cell pipette is used, draw the blood to the .5 mark, then very quickly draw the acetic acid solution to the 11 mark.  Rotate the pipette gently while the solution is being drawn into the bulb.  Close both ends of the pipette with thumb and finger and shake gently from side to side (not endways) to mix the blood with the solution sufficiently to prevent clotting.  Lay the tube aside, or give it to an assistant who will continue the shaking, gently from side to side.  If the tube is shaken vigorously the cells may be fragmented and if it is shaken endways the cells may be forced into the capillary tube, or the mixture of cells and solution may be forced into the dilatation above the bulb; if the contents of the bulb were well mixed the latter accident would be of negligible importance, but if part of the contents of the bulb were forced into the upper capillary tube of the pipette and the upper dilatation before the mixing is completed serious error would be present in the count.

    In our laboratories an erythrocyte pipette is used for the white cell count.  Draw the blood to the 1 mark on the pipette, then draw the acetic acid solution to the 101 mark; mix the contents of the bulb by gentle shaking in a sidewise direction, as already directed.

    The advantage of this greater dilution lies in the absence of the debris caused by the destruction of the red cells; they are all completely destroyed by the greater amount of acetic acid and completely dissolved in the greater amount of fluid.  The white cells are more easily recognizable.  In cases of leukemia and leucocytosis, this method is necessary for accuracy, and we think it preferable in all cases.

    Next, fill the red cell pipette for the red cell count.  Draw the blood to the .5 mark on the capillary tube, then draw the diluting fluid for red cells to the 101 mark.  Shake as directed for the white cell pipette.  Lay the tube aside until the count is to be made.

    The most frequent errors, in making the dilution, are these:  The end of the pipette may be allowed to reach the air, either by being raised too high or by being pushed through the drop of blood.  The entrance of air into the tube causes the “breaking” of the column of blood, or, if the air is admitted with the diluting fluid, bubbles are formed.  In either condition a fresh drop of blood must be taken in a clean pipette and greater care observed.  Sometimes the drop of blood is too small; if the blood has been secured with difficulty and if the column of blood reaches to the first mark below the .5, the count may be completed, and the correction made in the final computation.  If it is practicable to secure another drop of blood this is a more satisfactory method.  Sometimes the pipette is pressed against the skin too firmly, and the blood does not ascend into the tube no matter how hard the breath is drawn.  In such a case, the tip is apt to be slightly lifted, and the blood is apt to ascend suddenly through the tube and into the bulb.  When too great force is employed in the inspiration which draws the blood into the tube, the blood is apt to rush too rapidly upward, fill the tube, and sometimes the bulb itself.  The pipettes must be rinsed immediately else the blood may coagulate within them and cleaning becomes an extremely difficult matter.

    Take next the blood for differential counts and for such other tests as may be indicated.  The pipettes of diluted ordinary blood may rest for a time, even for an hour or two, without harm, but in cases of unusual fragility of the cells changes may occur more quickly.  It is best in all cases to make the actual counts as quickly after the blood is taken as is practicable.

    After the blood has been taken for the various tests indicated cleanse the ear lobe of the patient and note whether there is any indication of persistent bleeding.  The wound should be closed by this time, if it is not handled.  It may be kept open for five minutes or more by the slight manipulations necessary for wiping away the preceding drops of blood in preparation for further tests.
 
 

    Place the counting chamber on a perfectly level surface, and the cover-glass by the chamber.  Close the ends of the pipette containing the red cells and shake again, using a side-to-side staccato movement, rather gently, about one hundred times.  Avoid the endwise movement in shaking.  Remove the rubber tube and attach a stiff rubber bulb to the upper end of the tube, force out about four drops of the mixture and discard this.  This is in order that the diluting fluid in the capillary tube, which contains no cells or, at most, only a very few cells, may not be used for counting.  Let the fifth drop begin to form at the end of the capillary tube, and touch this to the surface of the counting chamber, near the rule area.  Avoid allowing the end of the pipette to touch the surface of the counting chamber; if the glass should touch the ruled area the markings might be scratched.  The amount of fluid necessary must be learned by experience; it is about half as much as would drop from the end of the capillary tube of the pipette if the pressure on the bulb should be increased.  Holding the cover glass by its edges, lower it slowly over the counting chamber; if a bubble of air happens to be caught in the fluid clean the counting chamber and take another drop from the pipette.  If the counting chamber itself is not filled with the fluid, clean the chamber and take another drop of fluid.  If the amount of fluid is too great, so that the moat is filled across, the cover glass cannot fit accurately and the count will be too high.  In this case clean the counting chamber and take another drop of the fluid. Only experience can win accuracy.

    If the counting chamber has an H-shaped moat, the procedure is somewhat less difficult. Place the cover glass on the slide first, and note that the cover glass fits the outer raised part of the chamber accurately.  The test of this accurate fitting is best made by means of the phenomenon known as “Newton’s rings.”  These are concentric bands of rainbow colors which appear when two glass surfaces are so closely in contact that the difference between their surfaces is not more than the distance measured by wave-lengths of light.  When the slide containing the counting chamber, covered by the perfectly plane cover glass, is held slantingly to the axis of vision, in changing position, these rainbow colors should be visible.  This means that the fitting is fairly accurate.  If the rings are not visible press firmly upon the cover glass; this may cause the necessary approximation of the two surfaces. If the rings are then visible, and remain so when the pressure is removed, the cover-glass fits accurately.  If the rings disappear when the pressure is removed, there is some dust or moisture present; clean the cover-glass and the counting chamber and repeat the process.  Occasionally a counting chamber is found upon which it is impossible to secure the rings.  Such a chamber may still be accurate so far as the counting is concerned.

    Another test for accuracy lies in the fact that two perfectly dry plane glass surfaces closely approximated adhere firmly.  Having placed the cover glass I position turn the slide over gently; if the cover-glass adheres, the fitting is probably good.  If the cover-glass falls off, the fitting is not good.  Cleanse and dry the cover-glass and the slide and try again.  This method of testing is less accurate than the finding of Newton’s rings but it is useful in the use of certain chambers.  A trace of moisture causes adhesion, separates the surfaces and seriously increases the count.

    If Newton’s rings cannot be produced on a certain counting chamber, the chamber may be tested by another in which these rings are produced.  Take the same pipette of blood and make counts of the red cells on both chambers, in one of which Newton’s rings have been produced, and in the other only the adhesion of the cover glass has been found well marked.  If the counts made on the two chambers agree within the limits of accuracy permissible for the method (not more than 3% of the total number of cells counted) then the second chamber is accurate enough for all clinical purposes and its use can e continued. If the two counts do not agree, the imperfect counting chamber should be returned to the maker.

    Having secured the proper fitting of cover glass and counting chamber, shake the red cell pipette as before using at least 100 vibrations. Discard the first four drops, and touch the fifth drop to the edge of the chamber at the edge of the slide.  The fluid runs beneath the cover glass and fills the ruled area at once, by capillary attraction.  If any fluid runs out and fills the moat, thus lifting the cover glass and increasing the depth of the counting chamber, clean the cover glass and the counting chamber and take another drop from the pipette, after shaking it as before.

    With both types of counting chamber, the later processes are alike.  Allow the filled chamber to rest for three minutes or so in order that the cells may settle to the bottom of the counting chamber.  The cells of normal blood settle             quickly; those of anemic blood and blood which is from persons with certain diseases settle slowly.  If any cells are still floating when the count is begun, this must be deferred for a time, until all the cells rest upon the bottom of the chamber.  Accurate counts are not possible when any cells are floating, because they are not in focus at the same level.

    Use a 1/6 objective and a one inch eyepiece on the microscope.  Find the ruled area and select a large square containing five rows of five each of the small squares.  Count all the cells in the upper row of five small squares and note the number found.  In this count include every cell which touches the upper line and the left hand line, and the intersection of these; also at the intersection of the right line and the upper line; exclude form the count all the cells which touch the right line and the lower line, the intersection of these, and the intersection of the lower line and the left line.

    Repeat for each of the five rows of small cells.  This makes a column consisting of five numbers, each of which indicates the number of cells found in five small squares.  The sum of this column indicates the number of cells in 25 small squares.  Select another area containing twenty-five small squares arranged in the same way, and repeat.  For ordinary work four such columns, indicating the number of cells in 100 small squares, is sufficiently accurate.  In cases of anemia, count the cells in 400 to 800 small squares, using two or more counting chambers of diluted blood.

    Compute as follows:  Multiply the total number of cells counted by the dilution, and this by 4,000 (which is the cubic content of each square in terms of a cubic millimeter, since each small square is 1/20 by 1/20 by 1/10 millimeter in size).  Divide by the number of square counted.  Example, 625 cells were found in 100 small squares, the blood being diluted 200 times.  The computation is:

    625 X 200 X 4,000 divided by 100=5,000,000 cells per cubic millimeter.

    Or, the average number of cells in each square is 6.25.  Each square is 1/4000 cubic millimeter in content, so that there must be 25,000 cells in each cubic millimeter of the diluted blood.  Since the blood was diluted 200 times, the number of cells I one cubic millimeter of undiluted blood is 5,000,000.

    The counting of the white cells follows a similar process.  The same tests for accuracy of fitting the cover glass and for placing the drop of fluid in the counting chamber are employed.

    If the white cell pipette was used, count the number of cells present in at least 800 small squares.  If the red cell pipette is used count the number of cells present in 4,000 small squares.  In either case the computation follows the same method,--total number of cells counted, multiplied by the dilution, multiplied by 4,000 and divided by the number of squares included in the count.  For example, if the white cells pipette has been used, with a dilution of 1:10, and if 300 cells were found while counting 800 squares, the computation is:

    300 X 10 X 4,000 divided by 15,000 white cells per cubic millimeter.

    If the red cell pipette was used, with a dilution of 1:100, and if 120 cells were counted within 4,000 small squares, the computation is:

    120 X 100 X 4,000 divided by 4,000= 12,000 ells per cubic millimeter of blood.

    Always count at least 100 cells, even if it is necessary to cover 16,000 or more small squares in the counting. In doubtful cases count 1,000 to 5,000 cells, using several pipettes if necessary.

    Many counting chambers have large squares around the area of rulings for the small squares; the area of these large squares is easily understood by following the lines which form the small squares outward.  By using these larger squares the white cell count is easily and quickly made, even though 8,000 small squares are necessary for accuracy.  The computations are made upon a basis of the small square in all cases.  This avoids any possibility of error due to the use of different units.

    This actual count gives the number of red cells and the number of white cells per cubic millimeter of blood.  Attempts have been made to substitute estimations of the blood cell volume for studies of the blood cell count, since it is often thought more important to know the total mass of hemoglobin-containing protoplasm than the manner in which this hemoglobin is arranged in cells. On the other hand, the manner in which the hemoglobin is divided into cells is an important factor in the oxygen carrying function of the hemoglobin, since hemoglobin arranged in a comparatively large mass with relatively small surface area is exposed to the air in the lungs less efficiently than an equal mass of hemoglobin divided up into smaller masse with relatively greater surface area.

    The hematocrit is a form of centrifuge in which small tubes are placed in opposite arms; each tube has 100 equal divisions.  The technique is simple:

    Secure a large drop of capillary blood by the method outlined for taking the blood for counting.  Place a rubber tube over one end of the glass tube; draw the glass tube perfectly full of blood.  Cover a finger with Vaseline, and cover the free end of the glass tube immediately.  Remove the rubber tube; place the glass tube in the arm of the hematocrit.  Place the other glass tube, filled with water, in the arm of the hematocrit, to balance the machine.  If the blood of two patients is to be centrifugalized at the same time, make marks with a grease pencil upon the outside of each glass tube for identification.  Start the centrifuge, gradually attaining the high speed within half a minute.  Stop the centrifuge and note the height of the column of red cells at one minute intervals.  When two successive examinations give identical findings, stop the centrifuge.  Each division of the glass tube represents 100,000 cells, if the cells are approximately normal in size and in hemoglobin content.  The column of cells is approximately equal to the column of plasma in normal blood.  The blood plasma can be used for the determination of bile and other pigments.  A very thin layer of fatty globules is occasionally seen at the upper end of the tube, in lipemia.

    In the anemias the number of cells in each division of the tube may vary very greatly, so that in cases in which accuracy of cell count is important the method has no value.  As a method of determining the actual volume of hemoglobin-carrying protoplasm the method is of value.

    The time required for complete settling of the red cells has been studied by many workers, and this has been shown to vary greatly in different diseases.
 
 

    New slides are greasy and must be well washed, first in warm soap solution, then in hot water.  A second soapy washing is often necessary.  They can be kept in acid bichromate solution made approximately as follows:  exact proportions are not necessary:

            Potassium bichromate                               10 grams
            Commercial sulphuric acid                        10 grams
            Distilled water                                           200 c.c.

    In routine work take the drops which flow after the blood has been taken for an actual count.  If only the differential count is to be made, prepare and prick the skin as directed for the actual count.

    Have ready microscope slides which are perfectly clean and perfectly dry, at least eight slides for each patient.  Touch one end of a slide to the top of a drop of blood, then touch this blood to the end of another slide.  Allow the drop of blood to flow along the angle between the two slides for a second or two.  Push the first slide along the surface of second slide, away from the drop of blood.  The blood follows the moving slide and leaves a thin, even smear upon the second slide.  Never push or pull the first slide along after the drop of blood, because thus many cells are injured and there is a tendency for some cells to cling to the first slide and thus to accumulate in groups.  Repeat this process for each slide.  If the amount of blood is scanty it may be necessary to take only six slides; if these cannot be secured prick the skin again.  Never take less than six slides of blood for a differential count.  If the condition suggests leukemia, severe anemia or the need for any special study it is best to take twenty slides or more.

    As the smears are made lay the slides, blood side up, upon a flat surface until they are perfectly dry.  Then put them into an envelope already marked with the name of the patient and the physician and the date and the hours of taking the blood.  These smears keep almost indefinitely and they can be stained by different methods for special study of different structures or inclusions.

    If the blood arranges itself in circular areas or rings, the slide was greasy.  If the blood forms stripes or bands, the motion was jerky.  It is necessary that the second slide be moved along the first in a steady, even, deliberate manner.  If there are threads of fibrin or small thick places in the smear, the blood was partly coagulated.  If the smear is too narrow and thick, the second slide was moved before the blood spread along the edge.  If the smear is too thick and spreads over the entire slide, the drop of blood was too large.  If it is too thin and spreads over too small an area, the blood drop was too small.

    If two or more specimens are under observation at the same time it is necessary to mark each slide for identification.  This is best done with a lead pencil.  Using a pencil with a rather soft lead, write an initial or an identifying number upon one end of the smear, near the end of the slide, before the slide is stained.  The pencil ruins the blood cells over which it passes and leaves a small amount of the lead; fixing the blood on the slide makes the lines as produced permanent.  By making the marks at the end of the slide they do not interfere with the counting, since this is done in the central area.
 
 

    Solutions required are:

    Eosin yellowish, 0.5 gram in 100 c.c. methyl alcohol. This fixes the blood and stains the acidophile structures.
    Methylene blue, 1.0 gram in 100 c.c. tap water, if the tap water is clearn and reasonably pure, or methylene blue,                         1.0 gram
    sodium bicarbonate                              0.1 gram
    sodium chloride                                    0.5 gram
    distilled water,                                   100 c.c.

    The methylene blue stains the nuclei and the basophilic structures of the protoplasm.  It stains also the malarial and certain other parasites.

    Take one of the slides already prepared and dried.  Place the slide on a level surface, smear side up, and drop upon it several drops of the eosin solution.  Let stand fifteen or a few more seconds; rinse gently in tap water.  Drop upon it a few drops of methylene blue solution, enough to cover the smear very abundantly; let stand a minute or a little more; add a few drops of tap water and allow to stand on the slide about two minutes, rinse with tap water.  Drain, allow to dry in air thoroughly, and examine, using oil immersion objective, and one inch eyepiece.  Or, after rinsing with water, mount in water under a thin cover glass and examine, using a dry one-tenth objective and one inch eyepiece.  The one-eighth objective does not magnify sufficiently for the finer details of cell structure to be visible.  For careful study of the cell structures, a one-eighteenth objective, oil immersion, is useful. In our laboratories the dry one-tenth objective is used for ordinary work and the one-eighteenth objective for careful study of selected cells in unusual cases.  See also “Other Staining Methods,” Page 325.

    Make a general survey of the smear in order to note the type of blood cells present.
 
 

    Begin the counting at one edge of the smear, move the slide, by means of the mechanical stage, so that the field is brought toward the observer (apparently) as far as the edge of the blood smear, and is carried as far to the right as the edge of the blood smear, or as far to the right as the limit of the mechanical stage permits.  Then move the slide toward the left, noting each cell and making the notation in the column devoted to that cell type.  When the slide cannot be moved further to the left, or when the limit of the smear in that direction has been reached, move the slide away from the eye the diameter of one field, so that some selected red cell which is barely in vision at the lower edge of the field is moved just beyond vision at the upper edge of the field.  Then move the slide toward the right, counting and listing the cells as they appear in successive fields.  Continue in this way, moving back and forth across the slide, until all the cells have been counted on the slide, or until the desired number of cells has been listed.

    With practice it becomes easy to carry the counts in mind, and to make the notations in groups of twenty, t en or five, as the case may be.  In our laboratories, neutrophiles are counted in groups of twenty cells, small hyalines I groups of ten, and other cells as units.  These habits are made for the sake of accuracy, convenience and speed of counting.  Each worker develops his own customs.

    Count until the total number of cells counted is at least five hundred, in cases which present no marked variations from the normal and which show no marked irregularity of distribution of the cells.  In abnormal cases at least one thousand cells should be counted, while in cases used for special study, in unusual cases, and in all the leukemias two thousand to twenty thousand cells or more should be counted.  In one of our cases of leukemia, with an actual count of 250,000 leucocytes of which 80% were myelocytes of different forms, it was necessary to make a differential count of 50,000 cells in order to secure satisfactory accuracy.

    The number of cells to be examined depends upon the fact that successive counts of any selected number give almost or quite identical results.  In normal blood successive counts of one hundred cells each give approximately identical figures for lymphocytes and for neutrophiles, but may give very different figures for eosinophiles, while basophiles may not be found at all.  Successive counts of 100 cells may give eosinophiles of ten per cent in one 100, and no eosinophiles at all for another 100.  (Eosinophiles have a tendency to be in groups even in the best smears of blood.)  Counts of successive five hundreds of approximately normal blood give satisfactory accuracy for such blood.

    In acute cases in which diagnosis must be made quickly, as in suspected pyogenic processes probably requiring speedy surgical interference, a differential count of two hundred cells may serve the necessary purpose and enable treatment to be initiated quickly.  But unless there is urgent need of haste, every count should include at least five hundred cells.  Even when this haste is imperative, the count of five to ten hundred cells should be carried on later in order that accurate findings may be kept on record for later study and comparisons.
 
 

    In another case, with an actual leucocyte count of 2,500 cells, the neutrophiles included one half, or 50% of the total count.  The differential count was based on 1,000 cells examined.  While making the differential count thirty-four megaloblasts and twenty normoblasts were noted.  That is, there were fifty normoblasts and eighty-five megaloblasts per cubic millimeter in this blood from a pernicious anemia patient, taken three days before his death.

    In this same way, the number of several other structures can be computed on the basis of the differential leucocyte count, and much useful information gained thereby.
 
 

    Total white cell count, 120,000
    Large hyaline                                                      .24%                   288 per cu.mm.
    Small hyaline                                                      97.00%         116,400 per cu.mm.
    Mononuclear neutrophiles                                  .31%                    372 per cu.mm.
    Polymorphonuclears                                           l.20%               1,440 per cu.mm.
    Eosinophiles                                                      .72%                   864 per cu.mm.
    Basophiles                                                         .53%                   636 per cu.mm.
 
    One megaloblast and three normoblasts were found in making the second st age of the count; these are too few to serve as a basis for accurate computation but they indicate that there is some beginning injury to the red bone marrow.  They would not have been found at all in making an ordinary differential count.  The figures thus secured are more nearly accurate than could be secured by making a differential count based on an examination of 30,000 cells using the ordinary technique and the time required for the counting was much less.

    In cases of marked neutrophilic leucocytosis and in cases of monocytic angina this two-stage method of counting is very much more accurate and more convenient than the ordinary method.
 
 

    The older method of staining with iodine-gum preparations has been superseded by the staining with the vapor of iodine.  A wide-mouthed, closely stoppered jar is kept for this purpose.  About one gram of iodine crystals is placed in this jar.

    Blood smears freshly made after the manner already described for the differential count are placed in the jar, smear side exposed to the vapor of the iodine, and allowed to remain for five hours or more.  The slides do not over-stain, and they may be left for several days without harm.  One hundred leucocytes should be examined, and if iodophilic granules are not found, the reaction is negative.

    Note whether the protoplasm of the white cells is diffusely stained and list such stained cells as iodophilic.  Note whether granules are free in the plasma or are within white cells; if so, whether they are most abundant within the hyaline cells or the granular cells.
 
 

    Have a sheet of paper with columns numbered from 1 to 5.  Rarely columns 6 and 7 will be required.  Begin at one edge of the smear, as in the differential count, noting the number of nuclei in each neutrophile, but disregarding all other blood cells.  Count in this way the nuclei in 100 neutrophiles.  If a cell contains two nuclei, place a mark under column 2, if it has four nuclei, place a mark under column 4, and so on, until 100 cells have been counted.  (Plate XIV).

    In counting the nuclei, a ring-shaped nucleus, even if slightly beaded in appearance, counts as a single nucleus.  If the nuclear masses are united by a band, they should be counted as one.  If they are united by a very thin filament of nuclear substance, they should be counted as two.  If any cell has its nuclei piled one above another, so that it is impossible to determine the number of nuclei within it, it may be passed without counting; but if more than two or three such cells are found, the count must be repeated, using a thinner smear, for the higher counts will be those most often passed, under such circumstances, and the findings will thus be lower than the correct number.

    Add each column.  The sum of the cells of column 1, plus twice the cells in column 2, plus three times the cells in column 3, plus four times the cells in column 4, plus five times the cells in column 5 and six times the cells in column 6, if any, equal the sum of the nuclei in 100 cells. This divided by 100 gives the average number of nuclei for each neutrophile.

    For example, in a certain specimen of blood there are:

        10 cells having 1 nucleus; or        10 nuclei in all;
        38 cells having 2 nuclei  ; or       76 nuclei;
        41 cells having 3 nuclei  ; or       123 nuclei;
         7 cells having 4 nuclei   ; or     28 nuclei;
         4 cells having 5 nuclei   ; or     20 nuclei.

    These 100 cells have, altogether, 157 nuclei; or an average of 2.57 nuclei per cell.

    The neutrophilic nuclear average of this blood is 2.57.  The nuclear average in   normal adult human blood is between 2.45 and 2.55.  In normal children the nuclear average varies from 2.00 to 2.4, according to age.
 
 

    Solution:  80 c.c. 100 alcohol
    20 c.c. water, triple distilled
    Warm gently to about 40 degrees C.
    Add: .2 gram alphanapthol (Merck)
        .15 methyl violet 5 B (Grubler)
        .2 c.c. hydrogen peroxide (must contain 3% of the gas)

    Uses dried blood films prepared as directed for the differential count.  The films should not be more than a few hours old.  Place the slide on a level surface and cover with six to eight drops of the solution.  Allow to stain and fix for half a minute.  Add an equal amount of distilled water and allow to stain for five minutes.  Rinse several times with water.  Cover the slide with basic fuchsin solution (0.01%) to counter-stain for 2 minutes.  Rinse with water; remove water with filter paper; dry in air; examine with oil immersion lens or mount in balsam.

    Basophilic elements, including nuclei, basophilic granules, basophilic protoplasm, erythrocytes and platelets take various shades of red and pink.  Eosinophile granules show a peculiar circular staining so that the granules look like rings.  Neutrophilic granules and the finer granules of the endothelial cells show bluish tints.  The difference between the neutrophiles and the endothelial cells lies in the characteristic nuclear structure and larger granules of the neutrophiles, and the characteristic nucleus and the smaller blue granules in the protoplasm of the endothelial cells.  The stain is useful for its purpose. Pappenheim’s solution is adapted especially to a study of nucleated red cells.

    Solution:--Take a saturated solution of methyl green.              30 c.c.
        Add saturated solution of pyronin.                                     10 c.c.
        This stain will keep for several days, in the dark.
        Fix smears with heat, avoiding excess.
        Flood slides with stain for five minutes.
        Wash in water, dry, examine with oil immersion.

    The nuclei of the normoblasts and nuclear fragments stain a clear blue, while basophilic granules within the red cells stain bright red.

    Ehrlich’s traced stain is now little used.  It consists of equal parts of saturated solutions of indulin, nigrosin and aurantia, mixed together after a difficult and tedious technique.  It can be purchased in powdered form.

    Ehrlich’s triple stain also is somewhat difficult to prepare.  It may be purchased ready made up, though the commercial preparations are not usually very successful.  Grubler’s stains are commonly used.  The solution is made as follows:
 
    Take a 100 c.c. graduate and measure the ingredients in the order given; do not rinse the graduate at all during the process.  As each substance is measured pour it into a 500 c.c. flask and shake vigorously for one or two minutes.

            Saturated aqueous solution orange G                                             13.0 c.c.
            Saturated aqueous solution acid fuchsin                                            7.0 c.c.
            Triple distilled water                                                                       15.0 c.c.
            Absolute alcohol                                                                             15.0 c.c.
            Saturated solution of methyl green (added drop by drop with frequent shaking of the flask)                                  17.5 c.c.
            Absolute alcohol, (added drop by drop, with frequent shaking of flask)                                                                10.0 c.c.
            Glycerin (added drop by drop, with frequent shaking of flask)                                                                              10.0 c.c.

    This mixture can be used at once, but it seems to improve during the few days following preparation.  It deteriorates within a few weeks, more rapidly in the light or if the bottle is shaken.

    To stain—Fix slides with heat and place on a level surface.  Cover with solution taken from about the center of the bottle containing the stain, using a glass pipette for the purpose.  Never shake the bottle.  Leave stain on slide for three minutes or more; the slides do not over-stain if left twenty minutes.  Rinse with water, remove excess water with filter or blotting paper, dry, examine with oil immersion or mount in balsam.

    Erythrocytes stain yellow or buff.  Normoblast nuclei stain a very dark green, almost black.  Nuclei of leucocytes stain dark green but not so deep a color as the normoblast nuclei.  Fine granules of the neutrophiles and the endothelial cells stain lilac or pale purple.  Coarse granules of these cells and of the eosinophiles stain crimson.  Basophilic granules do not stain.  This stain is useful for distinguishing certain types of granules but it is not useful for general work.  It is very difficult to secure good stains in cases of leukemia.  Occasionally a patient appears whose blood refuses to take the Ehrlich triple stain, for no perceptible reason.

    Leishman’s stain is best purchased in powder form.  For use make a solution of 150 mgs. of the powder in 100 c.c. of pure methyl alcohol.  Smears are best made on cover glasses for this stain.  Use the technique given for making smears on slides.  Place a cover glass, smear side down, in a watchglass.  Drop the stain into the watchglass until the cover glass floats.  Allow to fix and stain for three minutes.  Add an equal amount of distilled water, and allow to stain further for one minute.  Remove the coverglass and wash in water, drain on edge until dry, examine with oil immersion lens.  Or, mount in water and examine with dry one-tenth objective.  The stain gives a fairly good picture when freshly made.  After about ten days standing it gives a differential stain for the azur granules also.

    Red cells take a coppery tint; in polychromasia some cells are pinkish.  Nuclei are in shades of reddish purple or purplish red.  Cytoplasm is bluish or blue.  Eosinophile granules are coppery red.  Neutrophile granules take a pinkish color.  Basophile granules stain purplish or reddish purple.  Azur granules are cherry red.  The stain is fairly good for general differentiation.  Leishman’s stain has bee simplified and modified in many ways.

    Wright’s stain has been developed from Leishman’s stain.  Its preparation is rather difficult and the resulting powder not always successful.  The powdered stain, which is a precipitate formed by combining eosin yellowish with methylene blue under certain conditions, can best be purchased.  This powder is to be dissolved in methyl alcohol, 1.5 gm. powder to 100 c.c. methyl alcohol.  The solution keeps for a month or more.  Wright’s stain is useful for general differentiation.  The technique of staining is:

    Place the dried slide on a level surface.  Flood with the methyl alcohol solution, which fixes and stains the slide at the same time.  Allow the stain to stand on the slide one minute.  Add an equal amount of distilled water, and allow to stand for two or three minutes,--the longer period giving a deeper blue stain, but eosinophilic granules are more deeply stained in the shorter period of time.  Longer standing than three minutes may cause a precipitate to be formed.  Rinse in water for about half a minute.  The thinner areas should be pinkish or yellowish in tint.  Experience is necessary to determine the exact degree of differentiation which gives best results for each blood specimen.  Mount in water and examine by means of a dry one-tenth objective, or dry and examine by means of an oil immersion lens.  Red cells show pinkish or yellowish.  All nuclei are blue or purplish blue, varying in shading for different types of cells.  Neutrophile granules are pinkish or pale purplish in color.  Eosinophiles are brilliant reddish pink or cerise.  Hyaline cells show blue protoplasm which may be very dark or rather pale.  Platelets are blue or purplish.  Malarial parasites are blue with darker purplish, reddish or bright red chromatin.  Mast cells show deep blue granules.  The stain is fairly useful in general differentiation.  Tap water used for differentiation increases the blue tints and this is often desirable.

    Giemsa stain is much simpler than the ;methods described, and it gives excellent differentiation.  The formula is simple and the stain constant in quality.  The powdered stain may be purchased or it may be made up as follows:

                        Azur II eosin                            3.0 gms.
                        Azur II                                       .8 gm.
                        Methyl alcohol, c.p.,              375.  gms.

    Grind up the stains in the alcohol, using a small amount of the alcohol first  When thoroughly mixed add:

                        Glycerine, c.p.                      175  gms.

    The solution keeps for several months, and sometimes much longer.

    The technique of staining is simple also.

    Place slides on a level surface, flood with methyl alcohol for five minutes, drain but do not rinse.  Put fifteen drops of the stain on the slide, then add ten drops of distilled water; stain for fifteen minutes.  Rinse, drain, mount in water and examine s usual.  Or dry and examine with oil immersion lens.  They fade quickly in cedar oil; paraffin oil may be used instead.  One drop of half-saturated sodium carbonate increases the staining of the basophilic elements.  The use of tap water instead of distilled water gives better differentiation in our laboratories.
 
 
    PLATE XIV
        Determination of the neutrophile nuclear average.  Nuclei only are shown.
            I. Five single nuclei.
            II. Six double nuclei.
            III. Six triple nuclei.
            IV. Five quadruple nuclei.
            V.  Six nuclei of five lobes each.
            VI. Six nuclei of six lobes each.
 
 

    A.     100 c.c. distilled water

    5 drops saturated aqueous solution sodium hydroxide

    1 gram alpha-naphthol

    Boil this solution, cool, decant fluid from any residue which may be present.  Allow to stand three days or more before using.  This solution will keep a month or more.

    B.     100 c.c. distilled water.

    .5 gram basic paraphenylenediamine

    Mix without heat.  Allow to stand at least twenty-four hours before using.  This solution will keep a month or more.

    C.     100 c.c. distilled water

    5 c.c. formalin

    Technique of staining.  First mix together equal parts of A and B and filter.  This mixture must be used within an hour or so.

    Fix dried blood smear in solution C for five minutes.

    Stain with the mixture of A and B for five minutes.

    Rinse, mount in water and examine, using dry one-tenth lens, or dry and use oil immersion lens.

    Cells derived from lymphoid tissue show no granules.  Cells derived from bone marrow show blue granules.

    The reagents are difficult to secure and preparation of the solution is cumbersome.  It is rarely of value in diagnosis but it has given some good results in research work.

    Merck offers simpler reagents; beta-naphthol-sodium is sold in sealed glass ampoules as Mikrozidin.  Solution A is a 2% solution of Mikrozidin in distilled water.  He supplies also dimethylyaraphenylene-hydrochloride in similar ampoules.  Solution B is a 1% solution of this in distilled water.  Equal parts of the two solutions are mixed and the resulting greenish precipitate filtered off.  The further technique is the same as in the original method.  The oxidase granules appear brownish or blackish by this method, instead of blue; the significance is identical.
 
 

    To demonstrate the nerve endings in bone marrow it is best to use histological methods, for which see any text-book on histology.  The methods employed are too long for discussion in this chapter.