576. 8 . 0 7 8 . 39 : 576. 852 . 2 3

THE PHOSPHATASE REACTION AS AN AID IN THE C‘LASSIFICATION O F THE CORYNEBACTERIA

JOHN BRAY From the E.M.S. Pathological Service, London, Sector VI

(PLATES LXXII-LXXIV)

PRETIOVS work (Bray and King, 1943) has shown that C. diphtherict! produces no colour when grown for 24 hours on a phenolphthalein phosphate-containing medium, whereas many diphtheroids hydrolyse the compound with the production of varying degrees of colour on exposure to ammonia. These observations suggested that the phosphatase reaction might prove useful, not only as a means of differentiatsing C. diphtheria from diphtheroids but also as an aid in the classification of the Corynebacteria.

Barratt’s (1 924-25) classification based on fermentation reactions is the one generally accepted, but this grouping is unsatisfactory because, as Barratt pointed out, there is no correspondence between the cultural characters of the strains and their fermentation reactions. Rarratt did not describe the colonial forms of the diphtheroids, but as strains differing markedly in other cultural characters may ferment the same carbohydrates while strains of the same colonial form may show marked variability in their fermentation reactions, classifica- tion based on fermentation only would appear to be unsatisfactory. Here the colonial forms of the diphtheroids have been illustrated, the relation of these forms to the phosphatase reaction has been studied and a modification of Barratt’s grouping is proposed which is based primarily on colonial form and phosphatase activity with the fc3rmentation reactions iu an ancillary position.

SOURCES OW DIl’ITTHTCROIDS

Twelve strains representing 13arratt’s twelve groups were obtained from the National Collection of Type Cultures. Other strains were isolated in the routine work of the laboratory from genito-urinary, conjunctival, nasal, pharyngeal and cutaneous sources. Many diphtheroids were received through the courtesy of Dr Mary Barber and Dr M. Y. Young. All the strains from the latter source were derived from nasal and pharyngeal swabs from suspected or proved cases of dipht,lleria. In all, 222 strains of diphtheroids were examined.

XETHODS

The organisms were spread on 5 per cent. horse-blood agar and, after incubation, colonies were inoculated into Hiss’s serum-water sugar media and planted on Dorset’s egg-calcium phenolphthalein phosphate medium. TO

497

498 J . BRAY

secure the maximum reaction the organisms were planted thickly, several strains being tested on the same plate as described by Bray and King. Next day the phosphatase reactions and the appearance of the colony on blood agar were recorded, and the diphtheroids were subcultured for further study on Loffler’s serqm and Hoyle’s tellurite agar. Finally, in some cases, the morpho- logy of the bacteria was studied in films stained by Neisser’s and by Gram’s method.

Femnentntion tests With the exception of dextrin, the same range of carbohydrates was employed

as in Barratt’s investigation, namely glucose, maltose, mannitol, lactose, sucrose and galactose. The results were read on the 4th day and the tubes kept until the 10th day for further observations. The formation of acid and clot was regarded as the criterion of fermentation. Slight amounts of acidity, often noted in the lactose and galactose tubes, were ignored as leading otherwise to excessive complication of the fermentation formule The transient reactions with maltose recorded by Barratt were not a feature in our series except in the case of group I diphtheroids. The tendency of maltose to become hydrolysed on sterilisation (Mudge, 1917) may account for this discrepancy. In other respects the types of reaction agreed with those obtained by Barratt. Barratt’s groups VI and VIII (table IV) were not met with in our series. When the sugars fermented did not correspond with those anticipated from the appearance of the colony the tests were repeated.

Virulence tests

When the source and fermentation reactions suggested C. diphtherict? virulence tests were made. All strains from throat, ear or nose which failed t o ferment sucrose were tested, with the exception of strains of C . dip7htherict: isolated from cases of clinical diphtheria.

CLASSIFICATION OF DIPHTHEROIDS

Barratt pointed out that the type of growth of diphtheroids 011 agar varied from fine and moderetely fine to rather thick and very thick, and stated that the majority of the biochemical groups contained representatives of both fine and thick growths. To illustrate this point a fine-growing strain such as is shown in fig. 4 may give the same biochemical reactions as the thick-growing strain shown in fig. 6. As a rule thickness of growth on blood agar corresponds fairly

TABLE I

Relation between type of colony and phosphutusc reaction

I 1 Type of colony 1 Pkphatase

Large, white, smooth (L.W.S.) . Small, grey, smooth (S.G.S.) . Large, white, rough (L.W.R.) . Small, grey, rough (S.G.R.) .

+ or -

i + -

well with size of colony, small colonies being present when the growth is thin and vice versa. It was found convenient, before familiarity was gained in the recognition of the different colonial types, t o group

JOURNAL OF PATHOLOGY-VOL. LVI PLATX LXXII

DIPHTHEROID COLOSIES ON BLOOD AGAR, 3 DAYS' GROWTH. x 18.

FIG. 1.-Group Ib . (C. hofnmnnii). FIG. 2.-Group l a . Colony resembles 0. hofmannii.

PIC. 3.-Group Ic. Colony FIG. 4.-Group I I . Small, resembles U. hofrnannii. grey, smooth colony.

FIG. 5.--Croup III. L.W.R. colony showing daisy-head form.

FIG. 6.-Group I l l . Large, white, rough colony show- ing effuse edge.

PHOSPHATASE REACTIONS 03’ DIPHTHEROID BACILLI 499

the colonies of the non-pigmented diphtheroids into those which showed relatively large colonies, white or yellowish when confluent, and those which gave relatively small greyish colonies. A further subdivision of these into rough and smooth types was made (table I).

It was noted that the rough diphtheroids, both L.W.R. and S.G.R., were invariably phosphatase-positive ; the L.W.S. diphtheroids varied from strongly positive to negative ; the S.G.S. diphtheroids were invariably negative.

Diphtheroids forming large white smooth colonies on bbod agar

On blood agar the colonial forms of the L.W.S. strains are similar, being in most cases indistinguishable from those of C. hofmannii (fig. 1). Because of the variation in the phosphatase reaction, further criteria are required before subdivision of these diphtheroids is possible. However, a t least one type can be separated without difficulty ; it is distinguished by its profuse growth, greater elevation and larger size of colony (fig. 2) and a surface which is definitely matt or dull compared with the shiny surface of C. hofmannii. (This feature is not clearly brought out in the photograph.) The phosphatase reaction is invariably strongly positive and the fermentation of sugars is vigorous. The peculiar reaction with maltose-acid after one day and then rapidly returning to neutral-is characteristic of these strains. Twenty of them were examined and gave the fermenta- tion reactions corresponding to Barratt’s group I (table IV), that is, acid from glucose, sucrose and galactose. These strains, it is suggested, should form one group designated Group la.

The other strains showing C. hofmannii-like colonies can be separated into two groups, the first of which failed to ferment sugars (Barratt’s group XI). All the L.W.S. diphtheroid strains which were phosphatase-negative failed to ferment the sugars tested but the majority of non-fermenting strains were phosphatase-positive. Of 100 strains examined, 84 were phosphatase-positive and 16 negative. Of the positive strains, 12 gave a weak reaction, only brought up by prolonged alkalinisation. Thus 28 in all gave a weak or negative reaction. Although the non-fermenting group contains many examples of C. hofmannii a few diphtheroids of long slender morphology were included. A t present it is convenient to designate all these non- fermenting strains Crotbp Ib .

The second group of L.W.S. C. hofmannii-like diphtheroids (fig. 3) can be separated from the non-fermenting group by the regular fermentation’of glucose, maltose and sucrose (Barratt’s group VII), and also by the different type of growth on tellurite and egg media. On Hoyle’s medium, for example, the colony of C. hofmannii is grey, whereas that of these strains is jet black, and on egg these strains tend to form a moist liquid growth. They are also distinguished by

500 J . BRAY

their marked non-emulsifiability. All 10 strains of this group were phosphatase-positive and it is proposed that they should be designated Grotq Ic.

Diphtheroids forming snaall grey smooth colonies on blood ugur

Twenty-five strains of the thinly growing smooth type of Table I1 shows the variety of diphtheroid (fig. 4) were examined.

fermentations obtained. TABLE I1

The sugar reactions of the majority of strains resembled those of Barratt’s groups IS and VII (table IV). It is easy t o understand how, if fermentation reactions alone are relied on, confusion would arise between these two groups. All the strains were phosphatase- negative or reacted very weakly and gave the colonial appearance shown in fig. 4. 30th the N.C.T.C. strains representing Barratt’s groups IS and IV were of this type. Diphtheroids of this type are referred to by Gordon and Higginbottoni (1942) in their description of a medium for distinguishing between certain nasal diphtheroids and C. diphtherice intermedizis, which they closely resemble. Indeed, two of our strains had fermentation reactions identical with those of C. diphtherice. Three strains received from Dr Gordon were similar to those isolated by us.

It is suggested that all small grey smooth phosphatase-negative strains should be placed in one group designated Group II.

Diphtheroids forming large white rough colonies on blood agcir

There is a large group of diphtheroids showing a wide 3-ariation in the size and surface markings of their colonies, but having rough- ness as a common characteristic. All these rough strains, diatever their size of colony, were phospliatase-positive. The comiiion colonial forms are illustrated in figs. 5-8. Although, superficially, the colonies of the different strains appear quite different, the large ones all show a tendency to form an effuse spreading edge. This is particularly marked in the form illustrated in fig. 6. Some strains present a fraiikly daisy-head form (fig. 5 ) , others form thin flat colonies (fig. 8).

The variations in the fermentation reactions of the 40 rough diphtheroids examined are, shown in table 111.

PHOSPHATASE REACTIONS OF DIPHTHEROID BACILLI 501

Phosphatase Glucose Maltose Sucrose Galactose I--’ - - + + + 4- - +

+ + + i +

No. of strains

24 L.W.R., 8 S.G.R.

4 L.W.R., 2 S.G.R.

2 L.W.R.

The two N.C.T.C. strains representing Barratt’s groups I11 and V were both rough and phosphatase-positive. In discussing her group V Barratt remarks that her group I11 contains several strains in which the growth on agar and serum is indistinguishable from that characteristic of group V. It seems to us that all these rough phosphatase-positive diphtheroids showing a tendency to form an effuse, crenated edge or daisy head should be placed in one group designated Croup III.

Diphtheroids forming small grey rough colonies on blood agar Ten of these were examined and are illustrated in fig. 9. Colonies

of this type are common in cultures from the conjunctiva and uterine cervix, and the more thinly growing strains appear somewhat like rough variants of Barratt’s group I1 diphtheroids ; but not only have S.G.R. colonies been seen occasionally to appear on replating a L.W.R. strain, but also the reverse process, in which S.G.R. become trans- formed into L.W.R. colonies, is not uncommon. This transformation is illustrated in fig. 10. Barratt’s observations on certain group V strains were similar. Moreover, two L. W.R. strains cultivated anaerobically gave rise to a pure growth of S.G.R. colonies, which changed to L.W.R. on incubating aerobically. The majority of the fermentation reactions of these strains corresponded to Barratt’s group 111. It seems probable that the S.G.R. is a variant of the L.W.R. form and should be placed in Group III.

Pigmented strains Although no examples of Barratt’s groups VI and VIII were

isolated by us, the two N.C.T.C. strains representing these groups had similar large rough yellow daisy-head colonies. Both were phosphatase-positive, and, as Barratt says of these groups, only distinguished from each other by the fermentation of lactose by the group VI strain. If pigmentation is ignored, strains of this kind

JOURN OF PATH.-VOL. LVI 2 K

502 J. BRAY

would fall into the same group as the other rough strains, that is, into the proposed Group 111.

The remaining yellow diphtheroids which were isolated, 5 in number -3 smooth and 2 rough-were all phosphatase-negative and inactive towards carbohydrates. In none did the colony appearance suggest, a close relationship to C . hofmnnii, and it is difficult to place these strains in any of the proposed groups.

Unch&$ed strains

In any grouping of organisms there are always some strains which are hard to place ; Barratt’s type strains Va and X, for example, occupy a doubtful position. Both were L.W.S. and rather weakly phosphatase-positive. Type Va appeared to have lost its fermentative powers when tested by us and would therefore fall into group Ib. The growth on Hoyle’s medium somewhat resembled that of group Ic strains. Type X (a similar mannitol-fermenting organism was isolated by us) was the only representative of this group in Barratt’s series. It seems undesirable to found a separate group with only two representative strains.

Three other strains (fig. 11) were hard to classify as they were the only examples of L.W.R. diphtheroids which had a negative phos- phatase reaction. Biochemically, they fell into Barratt’s group IV. Morphologically they were minute, gram-positive, cocco-bacillary forms. There was no growth on Hoyle’s medium, and it is possible that these strains do dot belong to the Corynebacteria. Unclassified pigmented diphtheroids have already been referred to.

Bacterial morphology

While great variation was observed among individual strains from any of the groups, it can be stated that the diphtheroids which show wedge-shaped, clubbed, curved bacilli with irregular sides or Chinese letter formations usually belong to the rough groups; whereas in the smooth groups, straight sided, evenly shaped bacilli end palisade formation are common. Metachromatic granules are numerous in all except groups I1 and Ib, in which they are scanty. In groups I11 and Ic they are often exceptionally large and prominent.

Barratt ’s type strains

Table IV indicates the colonial form and phosphatase reactions of the N.C.T.C. strains. The fermentation reactions are those given by Barratt (1924-25). If these strains are representative of the forms most commonly occurring in each fermentative group it is difficult to see why, for groups VII and XI, the uncommon pigmented strains were chosen. The colonial form of Barratt’s types I, Va, VII and X

PLATE LXXTIT JOURNAL O F PATHOLOGY-VOL. LYJ

DIPHTHEROID COLONIES ON BLOOD AGAR, 3 DAYS’ (:I<OWTIt. Y 18

FIG. 7.-Group I I I . L.TV.R. FIG. 8.-Group 171. 12.W.R colony 3l10w111g I<’IG. $).--GIoup I l l . colony showing formation of thin flat type. Small, grey, rough pap1lla3. colony.

FIG. lO.-&oup I I I . Transformation of S.G.R. into L.W.R. type.

FIG. 11. ~ Unclassified, ( ?) cliyli- thcroid.

PHOSPHATASE REACTIONS OF DIPHTHEROID BACILLI 603

was L.W.S., of I1 and IV, S.G.S., and of I11 and V, L.W.R. Type VII showed a smooth yellow-pigmented and type XI a rough orange- pigmented colony. Type IX, of non-human origin, was rough, yellow-pigmented and phosphatase-positive.

TABLE IV Fermentation and phospliatase reactions of Barratt’s diphtheroids

1 I1 I11 Iv V Va VI VII VIII X X I

Appearance of colony

L.W.S. S.G.S.

L.W.R. S.G.S.

L.W.R. L.W.S. Pig. R. L.Y.S. Pig. R. L.W.S.

_____

Pig.. R. C. di, Ltherm

ilucose

+ + + + + + + + i- + + -

%altos(

f f + k + f + + + i-

T -

Galactose

Appearance of colonies. L.W.S. = laxge, white, smooth. S.G.S. = small, grey, smooth. L.W.R. = large, white, rough. L.Y.S. = large, yellow, smooth. Pig. R. = pigmented rough.

Fermentation reactions. + =acid production with or without

& in maltose column = transient acid

- = no acid production.

clot.

production.

Group IX, of non-human origin, is omitted.

Modijicution of Barratt’s grouping

The following table summarises the suggested modification of Barratt’s grouping, which involves the reduction of the number of groups from twelve to five.

TABLE V Grouping of diphtheroids based on colony f o rm and phosphatme reaction

I Suggested group I Barratt’s group 1 Type of colony 1 Phosphatase 1 reaction I I I -I - 1

Ia Ib Ic. I1 I11

I X I VII

I1 and IV 111, V, VI and VIII

L.W.S. ++ L.W.S. L.W.S. + S.G.S.

L.W.R. and S.C.R. I I I .~ -

This method of grouping is supported when comparison is made of the colour formed by the various groups on the phosphatase plate (fig. 12). When the growths are seen side by side the colours produced by Merent strains of the same group match closely, and it is often possible to place an organism in its appropriate group by observing the type of growth on the plate and the amount of colour formed.

504 J . BRAY

COLONIAL AND OTHER CHARACTERS O F PROPOSED DIPHTHEROID GROUPS

Group Ia. Blood agar (2 days), 2.0 mm., white, circular, convex, opaque, smooth, edge entire, digestion of blood marked. On coagulated serum, pinkish. Hoyle’s medium (24 hours), 1-0 mm. ; (2 days), 1.5 mm., jet black. Dorset’s egg, brownish discolouration. Ferments glucose, sucrose and galactose, reaction in maltose transient. Strongly phosphatase-positive. These strains are probably the same as “ B. hoagii ” or group A of Morse (1912).

Majority of strains identical with C. hofmannii. Hoyle’s medium (3 days), colonies remain grey, in contrast to those of groups Ia and Ic. Sugars not fermented. Moderately phosphatase-positive in 84 per cent. of strains. Corresponds to group D of Morse.

Group Ic. Blood agar (2 days), 1.0 rnm., white, circular, convex, opaque, smooth, entire edge, similar to C. h o f m n n i i , very difficult t o emulsify. Hoyle’s medium (2 days), 1.25 mm., jet black. Dorset’8 egg, tendency to moist liquid growth. Ferments glucose, maltose and sucrose. Phosphatase-positive.

Blood agar (2 days), 0.1 mm.; ( 3 days), 0-3 mm., smooth, structureless, transparent colonies, with entire edge ; (4 days), increase in opacity, slight central umbonation and concentric rings form. Hoyle’s medium (18 hours), minute, grey ; (2 days), 0.3 mm., black, or black-centred colonies with grey periphery, similar to C. diphtherice intermedius. Dorset’s egg, some strains show tendency to form moist liquid growth like group Ic diphtheroids. Fermentation of sugars irregular. Majority of strains ferment glucose, maltose and sucrose. Phosphatase-negative. May be group C or B. xerosis of Morse.

(Type a) Blood agar (2 daysj, 1.5 mm. (L.W.R.), smooth central boss or plateau and effuse edge (fig. 6 ; fig. 7 shows papillae beginning to form), Hoyle’s medium (24 hours), hinute grey ; (3 days), 0.75 mm., jet black, crenated edge. (Type b) L.W.R. daisy-head type (fig. 5 ) , some strains pigmented yellow. (Type c) Large flattened type. Blood agar (2 days), 3-4 mm., with exaggerated spreading edge (fig. 8). (Type d) Blood agar (2 days), minute rough craggy colonies. Hoyle’s medium (2 days), grey ; (3 days), jet black. Fermentation of sugars irregular. Majority of strains ferment glucose and maltose only. Strongly phosphatase-positive. All L.W.R. strains are probably group B or B. $avidus of Morse.

Group Ib.

Group 11.

Group 111. 4 colonial types, all rough.

PHOSPRATASE REACTIONS OF c. UIPHTEh’I<l& AND ATYPICAL TOXIQENIC STRAINS

As previously reported (Bray and King), 60 strains of G. diphtherim were tested for phosphatase and found negative ; of these, 26 were of the gravis type, 7 intemedius and 15 mitis. Nine were unclassified, but 3 of these proved to be virulent. The three N.C.T.C. strains 3985 (gravis), 3988 (intermedius) and 3990 (mitis) were also tested and found to be phosphatase-negative. The following strains of non- virulent C. diphtheria were also negative ; N.C.T.C. 322, 1047, 4950, 4951, 4952, and 4953.

All strains of C. diphthericc! grew moderately well on Dorset’s egg, though by no means as profusely as diphtheroids of groups Ia and 111. Much better growth was obtained than with diphtheroids of group 11. When C. diphtherim is incubated on the mediuni for 48 hours or longer a faint pink reaction is elicited, somewhat weaker than that produced by a weakly reacting C. hofmannii after 18 hours. Three strains of gravis, one of intermedizcs and two of mitis were

JOURNAL OF PATHOLOGY-Voz. LVI PLATE LXXIV

PHOSPHATASE RK4CTIONS O F DIPHTHEROID BACILLI

FIG. 12.-Phosphatase reactions of diphtheroid bacilli, numbered according to Yarratt's classification, and of G . diphtkerice (K) on calcium phenolphthalein phosphate egg medium after 18 hours' growth.

I Unrratt's gronping.

T XI

~ V I I 11 and T V

111, V, VT, and VII I

Ant lior's grouping.

IU Ib Ic I1 I11

PHOSPHATASE REACTIONS OF DIPHTHEROID BACILLI 805

tested after three days’ incubation. No significant difference in the small amount of colour formed by the different types was observed. This observation is not in accordance with the statement of Leibowitz and Avinery-Shapiro (1941) that “ whereas extracts of gravis contain an active phosphatase, this enzyme is absent in extracts of mitis.”

Atypical toxigenic strains. Toxigenic Corynebacteria which are related to both C. diphtherice and C. ovis have been described by Gilbert and Stewart (1926-27), Mair (1928), Petrie and McClean (1934) and Barratt (1933). Representative strains of these-N.C.T.C. 4887, 4888, 4889, 4890, 4891, 4892 and 4893-were tested on the phenol- phthalein phosphate medium but were not otherwise examined. All grew thickly and were strongly phosphatase-positive, thus differing from C. diphtheria, but resembling C . ovis N.C.T.C. 3846.

DISCUSSION

The antigenic relationships of the diphtheroids of human origin have been very little explored and no satisfactory serological classification of the group has been made. Attempts to determine the antigenic relationship between representatives of different groups have been made by Morse (1912) and Eberson (1918), but the identity of the strains was ill defined and only small numbers were examined. It is hoped that this study may be supported by further work on the antigenic relationships of the diphtheroids we have described ; mean- while it provides a basis for the grouping of the diphtheroids and their differentiation from C. diphtherice by means of the phosphatase reaction. At present the usefulness of the test is limited by the fact that so far no efficient substitute has been found for egg as a medium basis, although the use of salts of phenolphthalein phosphate more soluble than the calcium salt may render egg unnecessary.

Diphtheroids as a group are relatively weak hydrolysers of phenolphthalein phosphate in comparison with, for example, Str. pyogenes, and while a medium containing serum is satisfactory for the latter, a high proportion of diphtheroids, while growing well, fail to show a reaction on the serum medium.

The test seems particularly useful with diphtheroids of group 111. It has no value with those of group I1 and is only of limited value with the non-fermenting strains. When used as an adjunct to fermentation tests a number of unnecessary virulence tests on non-sucrose fermenters will be avoided. If the test is used alone it is possible to say, even if only a moderate degree of colour is formed, that the organism is not C. diphtherim, although the rare aberrant toxigenic strains such as those described by Barratt and others cannot thus be excluded.

For routine purposes it is suggested that tubes containing 5 C.C. of Dorset’s egg with 0.02 g. of calcium phenolphthalein phosphate should be used. To alkalinise the medium a plug soaked in ammonia

JOWN. OF PATE.-VOL. LVI 2 x 2

506 J . BRAY

should be pushed into the tube. It is possible to inoculate a suspected colony directly from a tellurite plate, but it is essential to be sure that no phosphatase-positive contaminants are present, otherwise misleading results will be obtained. The growth should always be examined microscopically and the bacterial morphology studied. Dorset’s egg emphasises the clubbing of C. diphtheria.

Extensive trials of phenolphthalein phosphate media for the differentiation of C. diphtherice in mixed cultures have not been made. One or two experiments have shown that weakly reacting organisms such as Monilia albicans and certain micrococci when growing as isolated colonies give a negative phosphatase reaction.

SUMMARY

1. By means of the phosphatase reaction it is possible to differentiate C. diphtherice from many of the common diphtheroids isolated froin human sources.

2. The significance of the phosphatase reaction in the taxonomy of these Corynebacteria is discussed.

I wish t o thank Sir Alexander Fleming for his interest in this work, Dr M. Y. Young, Dr Mary Barber and Dr M. Gordon for the supply of cultures, Mr G. F. Stevenson for assistance in the preparation of media, Mr V. Willmott for the photographs, Miss D. M. Barber for the coloured plate and Mrs E. Bray for sacrificing many eggs. I wish also to thank Dr E. J. King for advice and for the supply of phenolphthalein phosphate, which may now be obtained commercially from Messrs Hopkins & Williams, 16 and 17 St Cross Street, Hatton Garden, London, E.C. 1.

REFERENCES

BARRATT, K. M. . . . . . 1924-25. J . Hyg., xxiii, 241.

BRAY, J., -4ND KING, E. J. . . 1943. This Journul, lv, 315. EBERSON, F. . . . . . . 1918. J . Inf. Dis., xxiii, 1. GILBERT, R., AND STEWART, F. C. GORDON, M., 9ND HIGGINBOTTOM, 1942. This Journal, liv, 435.

,, . . . . . 1933. This Journal,xxxvi, 369.

1926-27. J . Lab. and Clin. Med., xii, 756.

b.

LEIBOWITZ, J., AND AVINERY- 1941. Nature, cxlvii, 745.

MAIR, W. , . . . . . . 1928. This Journal, xxxi, 136. MORSE, M. E. . . . . . . 1912. J. Inf. Db., xi, 253. MUDGE, C. S. . . . . . . 1917. J . Bact., ii, 403. PETRIE, G. F., AND MOCLEAN, D.

SHAPIRO, S.

1934. This Journal, xxxix, 635.