ON D I F F E ~ R E N T I A LSTAINING ATiEXANDER PETRUKKEVTTCH Osborti Zodloyzcul Laborntory, Y a l e University FOVR FIGURES I n studying cytoplasmic inclusions in the cells of the digestive system of spiders T was confronted with the problem of finding stains which would permit clear differentiation. The inclusions are all granular or globular, of intergrading sizes and of not less than six different organic substances, exclusive of fats and lipoids. Several kinds occur jointly in the same cells and cannot be easily separated either by their. size or shape. I n may cases they all stain alike or else exhibit a bewildering series of intergrading colors. This is especially true in the case of triple and quadruple staining such a s Xlillot 's ( '26) acid fuchsin, metanil yellow, light green, toluidin blue method which, as I am 11om quite certain, led him to wrong conclusions. To settle the question as to whether the intergrading colors were clue to differences in the chemical nature of the inclusions or to faults inherent in the staining methods themselves, an extensive analysis was found to be necessary. A great many stains and staining methods after difierent fixing fluids were tried with variable success. It m7us at this stage of my work that I had R conference with my friend Doctor Tolstoouhov, on the application of pH control in staining. Unfortunately, neither his methods, nor those of other investigators who used buffered solutions of stains gave the desired results. However, they put me on the track of certain phenomena involved in differential staining, which finally gave me the clue to a proper and reliable method. This method combining the principle of destaining with control of PIT promises to have such wide application in 267 T H E A N 4 T 0 3 1 1 C A L RECOKI), \OT.. 6 8 , NO. 3 268 .\LEShNDEG PETRUNIiEVITCH histology that it seertied desirable to publish it in siicli form its to make it possible f o r othei. investigators to a p p l ~it to ot h el* material. Let us first coiisider the factors involved in differential staining. In doing so we may disregard the question a s t o tlicl physical or cheniical nature of staining and use the ex1)ression ‘staining affinity ’ for the reaction which takes place wlieii a substance arid a stain a r e brought together. Diffcretitintion is a function of the staining affinity and may be due t o one or several of the following factors: The chemical c~mpositionand physical state of tlic substrata, their change uiidel. the influence of various fixing fluids, the isoelectric iwint, solubility in the staining and washing fluids, the chemical composition, purity ant1 dissociation coilstant of the stain, the pII of the stain, tlie concentration of the stain, cluration of staining, tho tclmpcrature at which the stain is used a i i d the respective properties of the washing fluids and niounting medium. 31aiiy of these factors are interrelated and have been extensively studied hy various investigators, making their further, detailed consideration liere unnecessary. A short list of references is given at the e i d of this paper. Further references will be found in the papers listed. F o r the better widerstanding of the method to which we may now turn oiir attention, the folloming points must be cmliliasi zed. 1. Fixation has a profound effect on staining. Thus the so-called refractive bodies i n tlie interstitial tissue of the diqestive orgaiis of spiders reniain colorless in acid fiichsin a t all pH values at room temperature after fixation in almost all fixing fluids geneidly ernployed, but stain a deep red after fixation in absolute alcohol or paranitroformol (4 parts of my cupric pai.ariitroy)lieiiol fixing- fluid with 1 part of 37 to 40r/c1reagent grade formalin, freshly added). 2. Duration of staining has an effect only when it is shorter than the time required for the completion of the staining reaction. From that time on duration has iio effect whatsoever. To achieve identical results one sliould, therefore, always briiig staining t o completion. DIFFEHEXTIAL STAINING 269 3. The solvent used in the preparation of the staining fluid has a distinct effect on staining even though not easily noticeable at a glance. Alcoholic and aqueous solutions of the same stain in equal eoncentration and adjusted to the same pH produce a different degree of diff’erentiation, aqueous solutions being distinctly more selective. Ruff er solutions used a s solvents for the stains show the same effect, The greatest differentiation is obtained by aqueous solutions adjusted to a certain pH by the addition of HC1 or NaOH. Nearest in effect come stains dissolved in acetate, phosphate and borate buffers. Citrate buffers a r e less suitable because they give a more diffuse staining. Stains dissolved in phthalate buffers give very poor differentiation and should be avoided. 4. The pI-1 value of the staining solution is of the highest importance in itself. Moreover, il is intimately related to the method of fixation. The same stain used after different fixing fluids may have the same effect, but more often has a different effect. I n the latter case, a t least after some fixing fluids, it may or may not be possible to achieve the same result by using the staining fluid at some other p H value. It is an old experience that staining after fixation in picric acid mixtures is difficult and unsatisfactory. The reason for this lies not in an inability of the tissues so fixed to become stained, but in the fact that they stain diffusely at all values of pH. 5. The concentration of the stain, when the stain is allowed to act to the completion of its reaction, has no noticeable effect on the result within very wide limits. A 0.017. solution, let us say, of toluitlin blue gives the same depth of staining a s a 0.1% or a 1%solution when used at the same pH to the completion of its reaction. When unbuffered aqueous solutions a r e used, the degree of concentlation may lead to a considerable change of the ~ € 1 .Thus a 2yj aqueous solution of acid fuchsiii showed a p H value of 2.85, while the same solution further diluted with distilled water to 0.257. had a p H 3.64. Griibler’s aurantia in a saturated aqueous solution had a pII 7.8. A freshly prepared solution of 0.1% had a pH THE A N A T O M I C 4 1 ~RKCOBII, VOI.. 6 8 , NO. 3 270 ALEXANDER PETRUNKEVITCH of ca. 6.4. It is clear that the superiority of staining in dilute unbuffered aqueous solutions as often recommended in preference to more concentrated solutions, is due to the change in the pH value and not to concentration as such. 6. The temperature at which a stain is used has little effect on the result within a wide range, provided the pH is kept constant. The latter, of course, changes with temperature. If the same pH is to be used at considerably different temperatures, the buffer solution has to be adjusted correspondingly. A t temperatures over 60°C. a profound change i n staining properties may be observed. Thus the above mentioned refractive bodies after fixation in Regaud remain colorless in acid fuchsin a t all pH values within usual room temperature limits, but stain a deep red at a temperature of 60°C. or more in the acid range up to about pH 4. 7. The pH of the washing fluid has a profound effect on the result, an effect which is different from that of the pH of the staining fluid. I n view of its importance in the application of the method proposed here, it will be discussed in greater detail below. To achieve uniform results the sources of error were eliminated by the following methods : a. Stock solutions of a certain concentration were made f roin samples of certified stains further purified whenever possible. Weighing was done on an analytical balance. Water was twice distilled in a Pyrex distilling apparatus. A volumetric flask was used for dilution. b. Measurements of pH were made to the secoiid decimal a t a temperature of 25°C. with the aid of a glass electrode and a n amplifier with a pliotron tube. c. Temperature was controlled by a n airbath of special construction, having a tolerance of O.O2"C., and a water bath with a tolerance better than O.l°C. d. Long series of 7 p sections were made from the same block and mounted without albumen, by stretching over warm distilled water in the usual way. The same species of spider was used in all experiments, namely Phormictopus DIFFERENTIAL STAINING 271 cancerides (Latreille), a native of Haiti. I n addition, some experiments were made with smears of human blood and sections through the intestine of the leopard frog and the newt. I t may be worth while mentioning that some of the cell inclusions in the spider material a re proteins and others carbohydrates. e. The sections were kept a t constant temperature in the staining solutions for 12 hours, although it was found that normal staining was attained i n about 1hour. f. F o r the study of the effect of p I I of the stain, the slides were washed directly in pure dioxan. Most stains are not o r very little soluble i n dioxan. It was found that even prolonged sojourn i n dioxan did not affect the staining density, but for convenience of handling the slides were transferred to xylene before mounting in dammar. g. I n the procedure of the staining method finally adopted, HCl, HC1 and distilled water mere used buffer solutions, for washing and destaining before dehydration with dioxan. The first step consists in the preparation of a series of staining fluids having the same concentration of stain at all values of ~ € 1 . It was found desirable to begin at the lowest pH a t which the stain does not precipitate and to make the series a t 0.5 pH intervals all the way up to pH 1 2 or to such a pH in the alkaline range at which the stain does not precipitate or become decolorized. The range is, of course, a different one for each stain. Some stains, such as eosin, are not soluble i n water at lower values of pH and as they are transformed into a corresponding acid, the latter was used in preference to the salt. Eosinic acid of a high degree of purity may be easily prepared from certified water soluble eosin Y (di-sodium eosinate). Dried and weighed, it can then be dissolved in absolute alcohol and used as a stock solution f o r the preparation of 0.1% solutions in 50% alcohol by adding the required volumes of HCl or NaOH and water. If desired, one may use aqueous solutions from about p H 5 to pH 12 by dissolving dried eosinic acid in graded solutions of NaOH. I t is advisable to measure the pH of each stock solution, as this facilitates a subsequent adjustment of the stain. A 2 272 ALEXANDER PETRUNIiEVITCH curve is made by plotting the volume of HC’1 and NaOH against the pH for a constant volume aiid concentration of the resulting staining solution. Once the curves have been prepared, staining solutions of any desired value of pH may be easily aiid quickly made from the graph. In cases of buffered stain solutions one should follow the same procedure because the stain often causes considerable change in the pH value of the buffer solution unless the concentration of the stain is very small. The next step is the plotting of curves f o r the density of staining of the various cell components a t all values of pH. I I I 1 Color comparators, unless specially designed and constructed for such work, a r e neither suitable, nor really necessary, because all that one wants t o know is the approximate re1n t’ive density in terms of maximum, medium, slight or zero, when staining is completely absent. Similar curves liave been published by other investigators and could be omitted here. However, a n interesting feature manifested itself when several curves were compared with each other. This is the presence of an optimum or greatest density of staining a t a certain pH, the curve droppiiig from this point in both directions (fig. 1). Another remarkable feature is that a curve for the same stain used after another fixing fluid may not 273 DIFFERENTIAL STAINING only be different in shape, but show a constant rise up t o pH 12 (fig. 2). Higher values of pH were not tried because of the deleterious effeot on tissues. An examination of the curves thus constructed, permits at a glance the selection of the pH at which greatest differentiation takes place. Thus f o r methylene blue after fixation in paranitroformol the differentiation is greatest between pH 2 and pH 3.3, because in this range neither the zymogen globules, nor the refractive bodies are stained. The last step is the selection of the proper pH value of the buffer solution to be used in destaining. It may be accepted I l l I I 1 I I P H 2 3 4 5 6 7 8 9 10 0 z Fig. 2 Methylene blue after Regaud. '1 as a general rule that the capacity of retaining a stain in a destaining fluid of a given pH is not of the same order a s the staining affinity at that pH. A few examples may illustrate this. Acid fuchsin, after fixation in paranitroformol, barely stains at pI-1 8 and does not stain at all a t pH 10. Maximum staining is at pI-1 2 when all cell organs and inclusions are almost equally dark red. At pH 7 the refractive bodies and zymogen globules are stained helow medium density, while nuclei and cytoplasm are of a very faint pink color. But if a section is stained at pH 2, or even a t pH 2.85, and then destained for half an hour or more in a buffer solution a t pH 7, all cell organs and inclusions loose their color completely, 274 ALEXANDER PETRUNKEVITCH except the zyrnogen globules, the refractive bodies and the plasmasomes of interstitial nuclei. They retain their dark red color. I n the case of light green the maximum density of staining is between pH 2 and pH 2.5, while no staining of any kind takes place at and above pH 8 after fixation in Regaud. But if a section is stained a t pH 2.4 and destained in a buffer solution at pH 8, the zymogen globules remain dark green, while all other cell organs and inclusions become completely decolorized. The same holds true for orange G, only that in this case the refractive bodies as well as the zymogen globules remain colored after fixation in Regaud. After fixation in paranitroformol, staining in orange G at pH 1.8 and washing in a buffer solution at pH 6 the refractive bodies alone remain colored, the rest of the structures losing the color completely. The analysis of the curves and the capacity of stain retention by certain structures at a pH at which staining is otherwise impossible open the way to correct double staining without intergrading colors. If a staining solution is used, containing a mixture of two stains, one is bound to get intergrading colors because both stains a r e of necessity used at the same pIi no matter what the value of the latter may be. Take for example Wright’s blood stain. If one adjusts i t to a certain pH, one gets without fail a beautiful double staining of human blood because of the great differences i n the isoelectric point and chemical composition of the various cellular structures. Even so, some of them show intergrading colors which do not appear after staining with separate solutions of eosin a n d methylene blue. Tf Wright’s stain is used for staining other material, such a s sections through digestive organs of spiders or vertebrates, one gets besides pure blue and pink all kinds of intermediate purplish colors. This is always the case when the staining optima of both stains coincide or lie close together and the staining affinity of the substratum is more or less of the same order for both stains. But we have seen that staining affinity and capacityto retain stain at certain values of pH a r e of different order. Herein DIFFERENTIAL STAINING 275 as in staining with single stain solutions, the way is open f o r differentiation without intergrading colors. The stains must be used separately, one after the other a t the pH of greatest density and washing after each stain done in fluids having a pH at which stain will be retained only in the desired cellular component. Thus in the case of the digestive organs of the spiders, if the sections are first stained in eosinic acid at pH 2.1, washed in 6 HC1, stained in methylcne blue a t pH 3.6 to 4 and rinsed in distilled water, only the zymogen globules in the enzyme cells are stained pink, refractive bodies remain colorless, while all the other cell components are stained in various degrees of density of pure blue color. Toluidin blue may be substituted for methylene blue a t the same pH with similar result. If it is desirable to retain blue color in nothing but basophilic granules, the second washing solution should have a lower pH. I n this case & HC1 or even 2 HC1 should be used. If acid fuchsin is used instead of eosinic acid, the sections should be stained in it at pH 2.8 to pH 3.5 and washed in a buffer solution at pH 7 o r 8 depending upon the fixing fluid, then counterstained with either mcthylene blue or toluidin blue as before. But this is not the only result which may be achieved by the method here described. An examination of the curves of some stains, such as eosin (fig.3), orange G (fig. 4) and aniline blue shows that they cross each other at certain values of pH. In the case of eosin all cell components are stained with ail equal degree of density. S t p H 3 zymogen globules are stained deepest, nuclei and cytoplasm less so and refractive bodies least of all. At p H 10.5 refractive bodies are stained deepest, zymogen globules less so and cytoplasm and nuclei least of all. This permits the use of the same stain, only at a different pH, f o r differentiation of other cell components. F o r example, if we stain the sections in eosin at p H 9 (instead of pH 2.15) wash in distilled water and counterstain with methylene blue or toluidin blue as before, the refractive bodies are stained pink, all the other small components presenting different shades of blue, iiicluding the zymogen globules which appear light blue. 276 ALEYANDEI'I PETRUNKEVITCH At first sight, the method appears to be a complicated and troublesome one. In reality its practical application is simple, reliable arid f w e from errors inhereiit in other methods. Moreover the results a r e strikingly clear aiid beautiful. One must remember that each curve is good only for staining after I I I I I I I I I I 7 8 9 10 It 0 z / * I 2 -_ Fig. 3 Ithsinic acid :iftt.r paranitroforniol. . Fig. 4 > 0r:rnge G a f t e r sul)liniatc forniol. the particular fixing fluid for which it was drawn arld possibly only for the material wliicli mas used. IIumaii or amphibian tissues may require different values of pH of !he staining fluids. But once a curve has been made it is not necessary to prepare stains or buEer solutions of the entire series of pH. The optimum value is all that is needed DIFFEREN’~1AL S T A I N I N G 277 for stains, while for washing HCl, HCl,’ buffer solutions of pH 3, 6, 7 and 8 and distilled water will meet all requirements. Selection of a proper pfI is done from examination of the curve. Staining may be safely done at room temperature for not less than 1hour, still better over night. Rinsing in distilled water is not only permissible, but fully warranted whenever the pH of the water (ca. 5.5) comes to lie between that of the stain and the destaining buffer solution. Moreover, such rinsing prevents rapid deterioration of buffer solutions. After final rinsing in distilled water the stained slides are dehydrated in pure dioxan and mountcd in dammar. This prevents loss i n staining density and leaves the color a s permanent a s the nature of the respective stain permits. Xylene may be used between dioxaii and dammar, if one so desires. Alcohol should be avoided under all circumstances and is not only undesirable on account of its destaining effect on many stains, but also, fortunately, quite unnecessary. F o r the convenience of those who may wish to t r y the method, but lack equipment for proper measurement of p H of colored solutions the following instructions for the preparation of staining fluids of certain pH values may be of use. It innst be remernbcred, however, that samples of even certified stains vary appreciably and that the resulting values of pH may dif€er from those given here. Acid fuchsin, 2% aqueous solution-pH 2.85. To make 50 cc. of a 0.2576 solution having a pH 2, take 25 cc. of a 0.5% solution, add 0.5 cc. of ,”, HC1 and dilute with water to 50 cc. Anilivze hlice, make a 0.5% stock solution. To prepare a 0.1% aqueous solution having a pH of ca. 7.4 dilute the stock solution with distilled water. To prepare a 0.1% solution having a pH 2, add to 10 cc. of the stock solution 6 cc. of HCl and dilute with water to 50 cc. Aurnizfia, 0.1% aqueous solution-pH ca. 6.4. To make a 0.1% solution having a pH 2.4, take 25 cc. of a 0.2% solution “Fixanal’ rengcntv of tlic E. de Haen Company, Germany, Pfaltz and Bsuer sole agents for U.S.A., were found t o be quite satisfactory. 278 ALEXANDER PETRUNKEVITCH of aurantia, add 3 cc. of $ HC1 and dilute with distilled water to 50 cc. Benxonmriizc, after paranitroformol fixation gives double staining a t pH 5.5 to 6.0. To get a solution having a p H 5.5 make a 0.1% solution of benzoazurine powder in pH 5.5 acetate buffer. Eosin yellowish in a 0.5% aqueous solution has a pH of ca. 9.5. Eosiizic acid (tetrabromfluorescic acid), stock solution 0.2% in absolute alcohol. To make a 0.1% solution i n 50% alcohol having a pH 3.3 take 25 cc. of the stock solution and dilute with distilled water to 50 cc. To make a similar solution having a pI3 2.6 take 25 cc. of stock solution, add 2 cc. of HC1 and dilute with water to 50 cc. To make a solution having a pH 2.1 take 25 cc. of stock solution, add 7 cc. of HC1 and dilute with water to 50 cc. F o r pH 2.25 use only 5 cc. of & HCl. Light green. To make a 0.1% solution i n 50% alcohol having a pH 2.4 take 10 cc. of 0.5% aqueous solution, add 25 cc. of absolute alcohol and 4 cc. of ,”, HC1 and dilute with water to 50 cc. Metaail yellow, 1%aqueous solution-ca. pH 8.0. To make a 0.1% solution having pH 2.4 take 25 cc. of a 0.2% aqueous ; HC1 and dilute with water to 50 cc. solution, add 3 cc. of This dye is soluble in dioxan in place of which chloroform should be used. Methylew blue, U.8.P.-aqueous stock solution 0.5%. To make a 0.25% solution having a pH 4.0 take 25 cc. of stock solution and dilute with water t o 50 cc. To make a solution having a pfI 3.0 take 25 cc. of stock solution, add 0.5 cc. of ,”, HC1 and dilute to 50 cc. To make a solution having a pH 2.0 take 25 cc. of stock solution, 5 cc. of -Z HCl and dilute to 50 cc. with water. Ora?zge G , 0.1% aqueous solution has a pH of ca. 8.4. To make a 0.1% solution having a pH 2.0 take 25 cc. of a 0.2% soliition, add 5 cc. of ,”, HC1 and dilute with water to 50 cc. Toluidim blue, a 0.1% aqueous solution has a pH ca. 4.3. h 0.1% solution i n 50% alcohol has a pH ca. 4.6. To make a x DIFFERENTIAL S T A I N I N G 279 0.1% aqueous solution having a pH 2.1 take 10 cc. of a 0.5% aqueous solution, add 4 cc. of ,”, HC1 and dilute with water to 50 cc. W r i g h t ’ s blood stain, to adjust the pH to 7.36 which gives the greatest differentiation with human blood smears make a 1%stock solution by dissolving the powder in pure methyl alcohol and add an equal volume of a phosphate buffer solution pH 6.3. Double s t a i n k g with eosin-methyle9ze blue in separate solutions to obtain greatest differentiation without intergrading colors. Pass the blood smear over a flame and fix it for 1 or 2 minutes in pure methyl alcohol. Stain in 0.1% eosinic acid in 50% alcohol, rinse in distilled water, stain in 0.1% solution of methylene blue, U.S.P. in acetate buffer pH 5.5, rinse in water and dry. For double staining with eosin-methylene blue of oxyntic cells in the stomach of vertebrates, fixed in Zenker’s fluid, sublimate formol or paranitroformol, stain the sections in 0.1% eosinic acid in 50% alcohol adjusted to p H 2.2 to 2.3 wash in approximately 0.01 normal hydrochloric acid, rinse in distilled water, stain in 0.1% solution of methylene blue, U.S.P. in acetate buffer pH 4, rinse in distilled water and dehydrate directly in dioxan. Do not expect to obtain the same result a t other values of pH. LITERATURE CITED BAKER,J. R. 1933 Cytological technique. London. BURKE,V. AND F. 0. GIBSON 1933 The Gram reaction and the electric charge of bacteria. J. Bacter., vol. 26, pp. 211-214. CONN, €1. J. 1936 Biological Stains. 3rd ed. Geneva, Kew York. DUTTON,L. 0. 1927 The status of Wright’s stain and some suggestions as t o its use. Stain Tech., vol. 2, pp. 105-108. 1928 Wright’s as a differential stain. Stain Tech., vol. 3, pp. 140-142. HATANO, SUKENISA AND S ~ I G E RIWATA U 1933 Hydrogen ion roncentration of histiocyte by vital staining with indicator dyes. J. Orient. Med., vol. 19, pp. 823-832, 980-986. HAYNES, R. 1927 Investigation of thiazin dyes as biological stains. 1. Stain Tech., vol. 2, pp. 8-16. 1928 Investigation of thiazin dyes on biological stains. 11. Stain Tech., vol. 3, pp. 131-139. 280 ALEXANDER I’ETlLTiNIiEVITCH HoLnrEs, W. C. 1929 The mechanism of staining. The case f o r the physical theories. Stain Tech., vol. 4, pp. 75-80. JEFFERS, K. R. 1934 Staining reactions of protoplasm and its fornicd cornponents. A cytological and biochernical study. J. Rlorpli., vol. 56, pp. 101-123. LASSET-R, PH., A. DUPAIX-L?SSFUR AND W.E’ALGEX 1933 Fixation dcs colorants par les 616ments microbiens. T. D6coloration de quelques solutions colorantes par les suspensions barthriennes. Influence du pII sur la niarche du ph6noniPne. Trav. Lab. M ivrohiol. Fac. Phnrni. N a n q , 1’. 6, pp. 50-54. LASSEUR, PII. A N D M. BFNOIT 1934 Observations sur la mettiode de Gram. Trav. Lab. Microbial. Fac. Pharm. Nancy, T. 7 , pp. 129-151. LISON,L. 1936 IIistiocheiriie aniniale. Paris. MILLOT, J. 1926 Contribution a l’histophj siologie des aran6ides. Suppl. Bull. Biol. France et Belg., T. 8, pp. 1-238. MILOVIDOV, P. F. 1936 On the distribution of tlrynionucleic acid in normal and pathologically modified cells ; also 011 the structure, cheniical and physico-chemiral properties of the cell nucleus. Investigations with the aid of the nucleic rcaction. Dissertation puhlishcd by ttie author in Russian. Pp. 1-312. Prague. NAYLOR, E. E. 1926 The hydrogen-ion concentration and the staining of sections of plant tissue. A41n.J. Botany, vol. 13, pp. 265-275. PIETTRE,31. 1933 Floculation, clans 1’organisme, des eolorants, colloides artificiels chimiquement definis. C. 11. Acad. Sci. Paris, T. 196, pp. 298-300. SFKI,121. 1932 Zur p1i)sikalischen Cliemie der histologischen F i r b u n g . Folia Anat. Japonica, vol. 10, pp. 621-654. 1934 Studicn iler electrischen Ladung und FBrbbarkeit der Erythrocyten. V. Bcstimmung des isoeleetrischeii Punktes der fixicrten Erythrocyten :ruf farberischem Wege. Zcitschr. Ges. Exp. hled., vol. 94, pp. 655-662. STEIRN,E. Th’. AND A. E. STEARN 1928 T h e rffect of the chrmical nature of a derolorizer on its functioning. Stain Tech., vol. 3, pp. 81-93. STRCGGLR, S. 1932 Ucher das Verhalten des pfiailzlichen Zrllkeriics gegeiiU1)cr Aiiilinfarbstoffen. Em Beitrag zur Methodik tler Bestiriiniung des isoclectrisclien Punkte.; tler Kernphasen. Zeitschr. Wiss. Biol., A M . E., E d . 18, R. 561-.570. Tor,sToouHov, A. V. 1925 Some practical :ipplications of the physicorhemica1 theory of differential stainiug : Blood, tissue and bacteri:t staining. 1Rktti)lcnc h l w cosin wntrr soluhle niixtures as a n unitersnl dye mixture. €’roc. N. Y. Path. Roc., N. S., vol. 25, pp. 147-159. 1928 The cffert of 1)reliniinnry trratment (fixing fluids) on staining properties of t h r tissues. Stain Trch., vol. 3, pp. 49-56. ~ _ 1929_ Detailed differcnti:ition of bacteria by means of a mixture of acid and hasic dyes :it different pI-1 values. Stain Tech., \ol. 4, pp. 87-59. I‘NNA,P . G. 1921 Clironiolyse. Abderlialtlen’s Hand. hiol. Arbeitsmetoden. Y a n r HA, G. AND T. ISIIII1933 TJcbcr die Wasserstoffionen Xonzentrazion und die isorlectrische Reaction der pflanzlichm Protoplasten, insbesondere dcs Zellkernes und der Plartiden. Protoplasma, Bd. 19, 8. 194-212.