The Sensitizing and Indicator Action of Victoria Blue and Janus Green on the Flocculation Reaction for Syphilis

Home , Janus Green B, Methyl blue, Methyl violet

In the case of sulphonamides, cultures resistant to sulphanilamide were resistant or partially resistant to the other members of this group with the exception of marfanil. Resistance once acquired seems to be permanent, and so far we have not been successful in reducing it in vitro. REFERENCES. ALBERT, A., FRANCIS, A. E., GARROD, L. P., AND LINNELL, W. H.-(1938) Brit. J. exp. Path., 19, 41. LANDY, M., LARKUM, N. W., OswALD, E. J., AND STREIGHTOFF, F.-(1943) Science, 97, 265. LEVADITI, C., AND MCINTOSH, J.-(1910) Bull. Soc. Path. exot., 3, 368. MACLEAN, I. H., ROGERS, K. B., AND FLEMING, A.-(1939) Lancet, i, 562. MACLEOD, C. M.-(1940) J. exp. Med., 72, 217. RAMMELKAMP, C. H., AND MAXON, T.-(1942) Proc. Soc. exp. Biol., N.Y., 51, 386. RUBBO, S. D., ALBERT, A., AND MAxWELL, M.-(1942) Brit. J. exp. Path., 23, 69. TILLETT, W. S., CAMBIER, M. J., AND HARRIS, W. H.-(1943) J. clin. Invest., 22, 249.

THE SENSITIZING AND INDICATOR ACTION OF VICTORIA BLUE AND JANUS GREEN ON THE FLOCCULATION REACTION FOR SYPHILIS.

F. M. BERGER. From the Public Health Laboratory, County Hall, Wakefield.

Received for publication November 9, 1943. DEAN (1937) found that isamine blue could act as an indicator of the reaction between an antigen and its homologous antibody. The addition of the dye to a mixture of horse serum and dilute antiserum produced a precipitate which was easily visible because it took up all the dye from the supernatant fluid. Prof. P. L. Suther- land suggested the possibility of using isamine blue as indicator in serological tests for syphilis. This dye did not prove entirely satisfactory, but certain other dyes were found which, when used together with suitable alcoholic heart extracts, caused in syphilitic sera not only the formation of coloured precipitates, taking up all the dye from the solution, but also very markedly increased the sensitivity of the reaction. The Effect of Various Dyes on the Flocculation Test for Syphilis. When investigating the action of various dyes on mixtures of alcoholic heart extracts with syphilitic sera, the following simple technique was used: To 10 volumes of Kahn's antigen, containing 0-6 per cent. of cholesterol, one volume of tinct. benz. co. B.P. was added. Of this stock antigen 0-2 ml. were pipetted to the bottom of a serological tube and a small loopful of powdered dye added. In a few seconds, after the dye had gone into solution, 2 ml. of 0-9 per cent. sodium chloride solution was blown rapidly into the antigen-dye mixture from a pipette. Suspensions obtained in this way were tested with syphilitic and normal sera immediately after preparation. Sera were heated to 56° C. for 30 minutes before testing. One drop of the antigen suspension was mixed on a slide with an equal 252 amount of serum. The slide was rocked for 4 minutes, and the fluid examined for the presence or absence of floccules with the unaided eye. Suspensions containing no dye did not show flocculation in syphilitic or normal sera. Suspensions containing varying but known amounts of dye were also prepared and tested with syphilitic sera of varying strength. In this way the relative activity of various dyes was studied. The behaviour of many dyes was also investigated by a " tube test." The antigen suspension containing a suitable dye was prepared exactly as for the slide test. Of this suspension 0.1 ml. was pipetted into a small serological tube and 0 1 ml. of serum added. The mixture was shaken in a Kahn shaking machine for 3 minutes and 0 5 ml. of saline solution was added. The results of the test were then read at once. With normal sera clear, transparent fluids were obtained. Syphilitic sera in the absence of dyes usually caused some flocculation, which manifested itself by the opalescent or turbid appearance of the fluid. In the presence of certain dyes this flocculation was intensified, causing the appearance of large floccules easily visible with the naked eye, and the dyes were adsorbed by the precipitate. In all 78 different dyes were examined. According to their behaviour in the tests just described the dyes could be divided into three groups. The first group comprised all those dyes which did not have any influence on the reaction. The intensity of

TABLE JI.-Cla8siftcation of Certain Dyes According to their Action on the Flocculation Reaction for Syphilis. Group I. Group II. Group III. No effect. Slight sensitization. No indi- Marked sensitization. Indi- cator action. cator effect. Alizarin. Acriflavine. Janus green B. Anilin blue w.s, Brilliant cresyl blue. Victoria blue 4R. Azolitmin. Brilliant green. Anilin blue spirit sol. Benzopurpurin. Chrysoidine. Isamine blue. Congo red. Crystal violet. Janus black. Eosin water sol. Dahlia. Methyl green. Erythrosin. Eosin alc. so]. Nile blue sulphate. Fuchsin, acid and basic. Gentian violet. Pyronine G. Haematoxylin. Janus red. Sudan black. Indigocarmine. Malachite green. Methyl red. Methyl blue. Naphthol green B. Methyl violet. Nigrosin. Methylene blue. Orange G and II. Neutral red. Orcin. Phenosafranin. Resazurine. Prussian blue. Scarlet R. Resorcin-fuchsin. Sudan III. Safranin. Sulphophthaleins (various). Victoria blue B. Trypan blue. Toluidine blue 0 Trypan red. the reaction was not increased in the presence of these dyes, and the dyes, when at all soluble in water, were not taken up by the precipitate. Sudan III, fuchsin, benzopurpurin and many other dyes belonged to this group. Dyes of the second group 253 sensitized the reaction. In their presence smaller amounts of syphilitic antibody sufficed to cause flocculation than in their absence. The sensitizing effect was not, however, very powerful, and relatively large amounts of dye had to be present for the effect to take place. Although the floccules which formed in syphilitic sera were coloured, the greater part of the dye remained in solution. Methylene blue, neutral red, crystal violet, acriflavine and many others belonged to this group, which can be briefly described as having a weak sensitizing and no indicator effect. The third group of dyes comprised those having both marked sensitizing and indicator effects. In the presence of these dyes heavy flocculation occurred rapidly, even with sera containing only little syphilitic antibody. The precipitate was intensely coloured and took up all the dye from the solution, leaving a water-clear supernatant fluid from which the whole of'the dye disappeared. This was in striking contrast to the tubes containing normal sera, which showed a uniformly coloured clear fluid. Janus green and Victoria blue appeared to be the best sensitizers. A few other dyes mentioned in Table I had a similar but not quite so marked effect. Isamine blue is only slightly soluble in alcohol, and due to this fact its indicator and sensitizing effects were bound to be weak. When, however, 'finely powdered isamine blue was suspended in the alcoholic heart extract, after dispersion a good indicator and sensitizing effect could be observed. These effects could not be obtained when similar amounts of dye were dissolved in the saline solution used for dispersion of the alcoholic extract and not in the extract itself.

The Action of Janus Green and Victoria Blue. To exert their indicator and sensitizing effects these dyes had to be dissolved in the alcoholic heart extract. As in the case of isamine blue, dissolving of the dyes' in the saline solution used for dispersion of the alcoholic extract did not cause any or only little sensitization. Addition of the dyes to mixtures of antigen with antibody did not cause sensitization of the reaction. In the absence of tinct. benz. co., Victoria blue and Janus green behaved differently. In the case of Janus green only the indicator effect of the dye was lost and the sensitizing action maintained, whereas in the case of Victoria blue both effects were lost. The relation between the amount of dye present in the antigen and the sensitivity of the test was also investigated. Fig. 1 shows that the sensitivity of the reaction increases with the amount of dye. It shows, further, that in the presence of relatively large amounts of dye the slide test was more sensitive than the tube test, whereas in the presence of small amounts of dye the reverse was true. The existence of the sensitizing action of Victoria blue and Janus green could be demonstrated in three different ways. Firstly, when alcoholic heart extracts wer-e dispersed in excessive amounts of saline solution, so that no reaction occurred in the presence of syphilitic sera, nevertheless powerful precipitation was observed in the presence of these dyes. Secondly, syphilitic sera containing amounts of syphilitic antibody too small to react with a given antigen without dye showed marked flocculation in the presence of Victoria blue or Janus green. Finally, these dyes sensitized the antigen-antibody complexes to the flocculating action of electrolytes, as in the presence of these dyes smaller amounts of electrolytes were sufficient for precipitation than in their absence. The fortifying action of Janus green was very much stronger than that of Victoria blue, as Janus green could not only intensify the action of cholesterin, but was also capable of taking over to a considerable extent the fortifying action of cholesterin. (Samples of dyes used in these experiments were examined for the presence of mineral 254 salts, particularly sodium chloride and sulphate, and these were not found to be present.) Table II illustrates the effects caused by Victoria blue and Janus green, and com- pares them with the fortifying action of cholesterin in the presence of two different concentrations of electrolytes. An ether pre-extracted alcoholic beef heart extract, to 10 parts of which 1 part of tinct. benz. co. B.P. was added, was used as antigen. The requisite quantities of cholesterin or dye or both were dissolved in portions of this antigen, as indicated in Table II. One volume of the various antigens was dispersed in 10 volumes of (a) 0 85 per cent. and (b) 2 per cent. sodiuin chloride solutions. These antigen suspensions were tested at once and at various periods of time after preparation with normal sera and with graded dilutions of syphilitic sera.

1:24 1:21 _ E/ ,1:18 K0 1:15

=1:12 / o 1:9 _ , 1-6~~~~- .21:6 - ; _ o- / o o Slide test 1:3 o--- Tube test 0 0 0-01 005 01 0-15 0-2 Amount of Victoria blue FiIG. L.-The influence of Victoria blue on the sensitivity of the flocculation reaction for syphilis. The syphilitic serum consisted of a mixture of 10 syphilitic sera and graded dilutions were prepared, using a sample of mixed normal serum as diluting fluid. The same set of dilutions was used throughout the experiment. All tests were carried out on the slide, and the results were read with the naked eye after 4 minutes rocking. The figures in Table II indicate the highest dilution of syphilitic serum giving a definite reaction with the respective antigen suspension. For example, figure 6 means that the respective antigen suspension gave a well-marked flocculation in a 1 in 6 dilution of syphilitic serum. The figure 0 means that no reaction was observed in undiluted syphilitic serum, and 0.5 signifies that only a doubtful reaction was obtained with undiluted syphilitic serum. Normal sera did not react with any of the antigen suspensions at any of the time intervals noted in Table 11. Table II shows that in the absence of both cholesterin and dye the antigen suspension did not show flocculation with syphilitic sera, not even in high concentrations of electrolytes. In the presence of suitable amounts of cholesterin and relatively large amounts of electrolytes a sensitive antigen suspension was obtained, provided that a sufficiently long time interval was allowed for ripening. In this case the sensitivity of the antigen was a function not only of the amount of electrolytes present, but also of the time interval which elapsed after addition. In the absence of cholesterin, 255 TABLE II.-The Sensitizing Action of Victoria Blue, Janus Green and Cholesterin in the Presence of 085 and 2X0 per cent. Sodium Chloride Solution. Sodium chloride contenit of dispersing fluiid.

0.850o 0.O. Cholesterin. Dye. Time) interval after dispersioni. Time interval after dispersioni. Antigen. °,. °. Houirs. Hours. 0. 1. 2.30. .. 8. 24. 48. 0. 1. 2.30. 5. 8. 24. 48. 06 B.02-5 . () O (0O O co 0 0 0 0 ) O

0 , . ( 6 .) - 5 1 1 . O 5I) :3 4 - 5) 4 *.6)t 9 =0 s . . \'.~V.13. 0-25 O 1) (t3 . ( ()t( O ( ) () Ot ( a J. (". -.. 0 6 6 6 6 25 6 6 6 6 6 c, . 0(I . V.B. 0-l . 9 9 9 !) 9 9 9 . 13 13 13 15) 15 135 13 0O . J.G(. 0 . 13 13 1 3 13 13 15 1 13 13 13 13 13 1I 15 V.B., VictoIria blue ; J.G., Janus green. Tho figures signify the highest dilution of a sample of syphilitic serum ill which flocculation occurred. For further explanation see text. Victoria blue and Janus green behaved differently. Victoria bluie possessed a cholesterin-like action in the presence of low electrolyte concentrations, i.e. it had some fortifving effect, which, however, needed time to make its full appearance. In the presence of 2 per cent. of sodium chloride (when cholesterin was not contained in the antigen) Victoria blue acted as protective colloid and not as sensitizer of the antigen suspension. Janus green, however, both in the presence of low and high sodium chloride concentrations, markedlv sensitized the antigen suspension, even in the absence of cholesterin. The remarkable feature of this sensitization was that it took place immediately after dispersion of the antigen, and that it then remained at a constant level for relativelv long periods of time. In the presence of cholesterin both Januis green and Tictoria blue possessed this peculiar effect of immediate sensitization to a still more marked extent. Under such conditions the sensitizing effect of Victoria blue was increased in the presence of higher electrolyte concentrations, whereas the sensitizing effect of Janus green was of the same order in the presence of both 0. 85 per cent. and 2 per cent. sodium clhloride concentrations.

The Nature of the Sensitizing Action of Dyes. The mechaniism of the sensitizing action of Janus green and Victoria bluie appears to differ from that of cholesterin. Cholesterinized antigens must be allowed to ripen in order to attain maximal sensitivity, whereas antigens sensitized by dyes possess maximal sensitivity immediatelv after dispersion. The fact that cholesterin intensifies the sensitizing action of dyes and vice versa also suggests different mechlanisms of action of the two sensitizers. Browning and Mackenzie (1924) brought forward evidence to show that the sensitizing action of cholesterin is due to its influence on the plhvsical state of the emulsion and to the operation of a chemical factor. Eagle (1930) believes tllat the coarser dispersion of the alcoholic antigen which takes place in the preseince of cholesterin is the sole cause of its sensitizing effect. He therefore assumed (Eagle, 1931) that otlher water-insoluble, alcohol-soluble substances would have the same effect. Water- insoluble, alcohol-soluble dves as Sudan III or scarlet R did not, hiowever, possess any sensitizing effect on syphilitic antigenis. whereas water-soluble Victoria blue alnd Janus green were good sensitizers. It appears probable that sensitization by dyes takes place as a result of a chemical combination of the reacting suibstance of the alcoholic hieart extract witlh the dve. In favour of this assumption is also the fact that both reacting substance and dye appear in the precipitate, and that one substance cannot be separated from the other by washing with saline solution. One of the distinctive features of the sensitizing action of dyes is the fact that in their presence much smaller concentrations of electrolytes are sufficient for precipitation than in their absence. The mechanisin of the sensitizing action of dyes, however, differs from the flocculating action of electrolytes. Sensitization of negativelv charged antigen suspensions could be obtained not only by the positively charged Victoria blue or Janus green, but also by the negatively charged isamine blue. A further difference between the action of electrolytes and that of dyes is the fact that the former can sensitize the dispersed antigen suspensions, whereas the latter have to be present before dispersion for sensitization to take place. There is a point of similarity between the sensitizing action of dyes and the action of antibody on antigen. Both antigens sensitized by dyes and antigens sensitized by homologous antibody possess an increased sensitivity to the flocculating action of electrolytes. The sensitizing action of the dye may be pictured as an antibody- like effect which has been deprived of its specificity, and which iA conditioned by the presence of a specific factor in syphilitic serum. Dean (1937) also observed that isamine blue precipitates globulin under suitable conditions, and that the process resembles an antigen-antibody reaction in respect of optimum proportions. Marrack (1940), when studying precipitation of horse serum euglobulin by isamine blue, also concluded that the reaction resembles the precipitation of antibody to pneumococci by the specific polysaccharide in several respects. The study of similar systems, as has been already pointed out by Marrack, and particularly the study of the mechanism of the sensitizing action of certain dyes, may help to elucidate the reaction between antigen and antibody.

The Dye Test for Syphilis. Making use of the sensitizing action of certain dyes, a practical test for the serological diagnosis of syphilis was devised. Sensitization by dyes gave the test peculiar advantages of its own, namely, the ease and exactitude with which the sensitivity and reliability of the test could be standardized and maintained at a desired level. The sensitivity and specificity of an antigen for serological tests for syphilis, when not containing dyes, is influenced very markedly by ageing or ripening, i.e. the time allowed to pass between dispersion of the antigen and its use in the test. Standardization of the sensitivity of the test becomes difficult, as ageing is a continuous process. In the presence of a suitable dye, however, the test can not only be adjusted to a desired sensitivity by alterations in the dye content of the antigen, but the antigen suspension can also be rendered more stable. It will possess its full sensitivity immediately after dispersion, and the same level of sensitivity will be maintained for some time thereafter. Very fine dispersion of the alcoholic antigen was obtained in the presence of Victoria blue. Therefore this dye was used for sensitization in the test in preference of Janus green. Many different alcoholic heart extracts and fractions therefrom were found equally amenable to sensitization by dyes. Kahn's extract is recommended because it is easily prepared and also obtainable commercially. Preparation of the stock antigen.-To 10 parts of Kahn's standard antigen containing 0-6 per cent. of cholesterol and having a titre of 1 + 1-2, add 1 part of tinct. benz. co. B.P. To 10 ml. of this mixture add 17 mg. of Victoria blue 4R (Gurr). This stock antigen keeps at least for a year and probablv indefinitely. It must be kept well stoppered, and care taken to prevent evaporation during preparation and handling. 257 Should an antigen of greater or lesser sensitivity be desired, the amount of dye may be increased or decreased. The use of antigens of greater sensitivity is, however, not recommended, as further increase of sensitivity is not possible without a decrease in the specificity of the reaction. Dispersion of the antigen.-One part of the stock antigen, say, 0.1 ml., was pipetted to the bottom of a serological tube, and 10 parts (1 ml.) of 0-85 per cent. sodium chloride solution were blown rapidly into the-tube from a pipette having a large open- ing at the tip. One ml. of dispersed antigen was suffipient for the performance of at least 30 tests. The antigen suspension could be used at once after dispersion, or at any time on the day of dispersion. It should be shaken immediately before use. Teckni4que of the test.-Sera to be tested were heated in a water bath at 56° C. for 30 minutes or at 620 C. for 4 minutes. One drop of heated serum was put on a glass slide and a drop of equal size of the dispersed antigen added. The drops were mixed with a glass rod and the slide rocked for 4 minutes. For convenience a large 4 by 6 in. glass slide, which was divided into 20 compartments by glass ink, was used. To read the results of the test the slide was placed on a sheet of white paper, and if the reaction was positive, large, deep blue floccules were easily visible in a water-clear fluid. If the reaction was negative the drop remained homogeneous and of sky blue colour. To detect the presence of small blue clumps in sera giving doubtful results, one end of the slide was lifted an inch or so above the paper. The results were read with the naked eye. Comparison of the dye test with the Wassermann and Kahn tests. The sensitivity and specificity of the dye test was compared with that of Wyler's (1929) modification of the Wassermann reaction and Kahn's (1928) standard test on 1112 sera. Table III gives a summary of the results of the three tests. In all, 187 positive and doubtful sera were detected by the combined use of the three tests. The table shows that the dye test was of greater sensitivity than either of the other tests. There were 26 sera reacting in the dye test but giving a negative result in the Wasser- mann and Kahn tests. The clinical histories of the patients whose -sera reacted only in the dye test were investigated, and it was found that 5 were cases of early untreated syphilis, 3 cases of late untreated syphilis, 15 were cases which were under treatment for syphilis or had treatment for the disease at a previous date, 1 was a case ofgeneral paralysis of the insane, and in 2 cases information as to the clinical condition of the patient could not be obtained. The result of this inquiry showed that the results given by the dye test were specific.

TALBE III. -Summary of Results Obtained with 1112 Sera Examined by the Wassermann, Kahn and Dye Tests. Test. Number of positive Number of doubtful Number of negative Test. results. results. results. Wassermann . 131 . 20 961 Kahn . . 124 . 19 . 969 Dye . . 160 . 18 . 934 Although the dye test is somewhat more sensitive than Wyler's modification of the Wassermann test or Kahn's flocculation test, it was not as sensitive as the collodion particles agglutination test (Berger, 1943). The dye test is, however, easier to perform, as very little laboratory equipment is required for its performance and the stock antigen keeps indefinitely. 258 DISCUSSION. Addition of various dyes to syphilitic antigens has been recommended by numerous authors. The sensitizing action of certain of these dyes seems, however, to have been overlooked. Borowskaja (1927-28) recommended addition of Victoria blue to the antigen for Meinicke's turbidity test in order to facilitate reading of the results. She noted the indicator effect of the dye, but did not notice its sensitizing effect. Later, Meinicke (1934) added 0-01 g. of Victoria blue to 100 ml. of his antigen for the clari- fication test. Victoria blue in that concentration could not have exerted any sensitization, and Meinicke did not give his reason for the addition. Rosenthal (1929-30) added methylene blue to his antigen to make the final results more conspicuous for reading, and Marquez (1941) added acriflavine for similar reasons. Laughlen (1935), Levy (1941), Leiboff (1940), Lorenz (1940) and Rappaport and Eichhorn (1943) added fat-soluble Sudan III or scarlet R to their respective antigens. Ide and Ide (1936) coloured their antigen with crystal violet and azur II. Sensitization by dyes does not seem to be confined to reactions for syphilis. It may be possible to find suitable dyes which would sensitize various antigen-antibody reactions. A few observations pointing in this direction have been recorded in the literature. Dean (1937) pointed out that addition of isamine blue to mixtures of horse serum and its antiserum produced a precipitate which was larger than that in a com- parable mixture of antigen and antibody without dye. Dean also found that the use of isamine blue may extend the range of dilutions in which an antiserum could be used. Berger and Brecher (1939), when describing a slide test for the exclusion of typhoid fever, found that the addition of crystal violet to the suspension was of some importance for the sensitivity of the reaction. Other dyes which were tried did not possess the sensitizing action. Boyd (1939) observed that antisera for blood group factors, when treated with acriflavine or brilliant green, kept better than those simply kept sterile or treated with toluene. The dyes may have acted as sensitizer in these cases.

SUMMARY. 1. The influence of a number of different dyestuffs on the flocculation test for syphilis was investigated. 2. Janus green, Victoria blue and certain other dyes had both a sensitizing and indicator action on the reaction. Syphilitic sera containing too little of the reacting substance to react with a suitable heart extract without dye showed marked flocculation in the presence of Victoria blue and Janus green. These dyes were also taken up by the precipitate, leaving a water-clear supernatant fluid. 3. The sensitivity of the reaction was a function of the amount of dye. Antigens containing dyes attained their full sensitivity immediately after dispersion, and maintained the same level of sensitivity for some time thereafter. 4. Making use of the sensitizing effect of Victoria blue, a simple and reliable slide flocculation test for the serological diagnosis of syphilis was devised. This dye test was just as specific as, and somewhat more sensitive than, other commonly used serological tests for syphilis. The results of the test were easy to read. 5. The mechanism of the sensitizing action of dyes is discussed, and the possibility of sensitization of other antigen-antibody reactions by suitable dyes is pointed out. This investigation was carried out at the instigation of Prof. P. L. Sutherland, to whom the author is indebted for much valuable criticism. The author also wishes to thank Prof. H. R. Dean for a sample of isamine blue. 259 REFERENCES. BERGER, F. M.-(1943) J. Path. Bact., 55, 363. Idem AND BRECHER, G.-(1939) J. Hyg. Camb., 39, 294. BOROWSKAJA, D. (1927-28) Zbl. Bakt., 105, 455. BOYD, W. C.-(1939) J. Immunol., 37, 65. BROWNING, C. H., AND MACKENZIE, I.-(1924) 'Recent Methods in the Diagnosis and Treat- ment of Syphilis.' London, 2nd ed. DEAN, H. R.-(1937) J. Path. Bact., 45, 745. EAGLE, H.-(1930) J. exp. Med., 52, 747.-(1931) Ibid., 53, 605. IDE, S., and IDE, T.-(1936) J. Lab. clin. Med., 21, 1190. KAHN, R. H.-(1928) 'The Newer Knowledge of Bacteriology and Immunology,' ed. by E. 0. Jordan and I. S. Falk, Chicago, 848. LAUGHEEN, G. F.-(1935) Canad. med. Ass. J., 33, 179. LEIBOFF, S. L.-(1940) J. Lab. clin. Med., 25, 317. LEVY, M. A.-(1941) Am'. J. clin. Path., 11 (Technical Section), 150. LORENZ, W. F.-(1940) Ibid., 10, 527. MARQUEZ, F.-(1941) Am. J. Syph., Gon. and Ven. Dis.; 25, 319. MARRACK, J. R.-(1940) Proc. III Intern. Congr. Microbiol., New York, 821. MEINICKE, E.-(1934) J. Lab. clin. Med., 19, 518. RAPPAPORT, F., AND EIcHHoRN, F.-(1943) Lancet, i, 426. ROSENTHAL, L.-(1929-30) Proc. Soc. exp. Biol. N.Y., 27, 61. WYLER, E. J.-(1929) 'Medical Research Council, Spec. Rep. Ser. no. 129.' London.

260