REMARKS ON TYPHOID BACILLURk4: WITH A DESCRIPTION O F CERTAIN ATYPICAL COLI- TYPHOID BACILLI FOUND IN THE URINE IN ENTERIC FEVER.‘

By ADAM PATRICK, M.A., M.D., Assistant to the Professor of Practice of Medicine, Glasgow Univemity ; late Resident Assistant Physician, City of Blasgow Feve?. Hospital, Ruclzill.

From the Luboratory of the Fever Hospital, Xuchill, Clasgow.

IN the course of a routine examination of the agglutination properties of typhoid bacilli, isolated from a number of patients with enteric fever, it became apparent that an important fact with regard to ‘: typhoid ” bacilluria has hitherto been overlooked, namely, that the bacilli which appear in the urine, and which apparently have always been classified as B. typhosus, are not always that organism. In six out of seventeen unselected cases of bacilluria in enteric fever it was found that the bacilli were atypical members of the typhoid-coli group. With the exception of organisms occasionally present as contaminations, the bacilli in the urine in enteric fever have always been proved to be, or regarded as, typhoid bacilli.

The facts with regard to typhoid bacilluria, as recorded in the literature, are the following.

I n a certain number of typhoid patients (about 25 per cent.) large numbers of bacilli suddenly appear in the urine, usually late in the course of pyrexia, or in early convalescence. Their presence is easily detected by the so-called opalescence, or shimmer, of the urine, and by the formation of visible waves in the depths of the fluid on shaking. As a general rule they are present in large numbers-172,000,000 per C.C. (Petruschky, 1898 I), 500,000,000 per C.C. (Gwyn, 18992)-but they have been found also in a few cases i n small numbers in apparently clear urines (Connell, 1909 ; Euchan, 19084). Bacilluria is commonly unsssociated with any special symptom, and, as a rule also, is unassociated with the occurrence of pus or albumin in the urine.

The bacilli persist in the urine for a variable period, often for several Received October 23, 1913.

366 A DAM P A TUCK,

weeks, and then commonly disappear spontaneously ; but in some instances they last for years,-nine years in a case recorded by Liebetrau (1906 5 ) , five years in it case mentioned by Gwyn, and in cases under the supervision of the Glasgow Public Health authorities for four to six years.

Connell collected the cases from the literature, and found, between 1897 and 1909, 150 instances of bacilluria in 621 cases of enteric fever. I n all these cases the bacilli were proved bacteriologically to be typhoid bacilli. To this list may be added seventeen instances of bacilluria in thirty cases (Buchan), and twenty-six in 100 cases of enteric fever studied by M'Call in the City of Glasgow Fever Hospital, Belvidere. These figures show that bacilluria occurred iii 25 per cent. of 751 cases of enteric fever.

The cause of typhoid bacilluria has been variously regarded. The fact that it comes on comparatively late in the disease, while bacilli are present in the blood in greatest numbers in the early stages, makes it unlikely that they are excreted directly from the blood.

Horton Smith (1900 6 ) found the blood sterile in four cases of bacilluria, and Coniiell in two cases. Konjajeff (1889 7) held that bacteriuria indicated the presence of lymphoid nodules in the kidney, for in sections of these nodules he had sometimes seen bacilli present. Suppurative foci in the kidney have been described by Flexner (18968), and by Erownlee and Chapman (1906 9). These small kidney abscesses, however, are not very common, for in 289 post-mortem examinations, Horton Smith found them onee only, though they have occurred more conimonly than this in the ex- perience of the City of Glasgow Hospitals. Hlumer (1895 lo) was of opinion that the urine was infected by direct passage of bacilli from the rectum to the bladder.

That the bacilli are sometimes confined to the bladder was shown by Horton Smith, who, at the post-mortem examination of a patient in whom bacilluria had been present, isolated B. typliosus from the bladder, and not from the kidney ; and by Gwyn, who caused bacilluria to terminate in three cases by washing out the bladder with a weak perchloride of niercury solution.

The commonly accepted view of the etiology of bacilluria is that the urine becomes infected by a stray bacillus a t an early stage in the disease, while organisms are numerous in the circulating blood, and that the bacilli multiply in the urine later, when its reaction becomes favourable. During the earlier part of au attack of enteric fever the urine is quite acid, and typhoid bacilli grow best in faintly acid media. Later in pyrexia, or in early convalescence, the urine loses its high acidity, and thus becomes a suitable culture medium. The presence of residual urine has been looked on as of importance in harbouring germs.

Park (1901 11), Horton Smith, and others made experiments on the growth of typhoid bacilli in urine. Buchan noted that the bacilli would not grow if the acidity of the urine rose above N/40, but he found them present, though not multiplying, in urine of acidity as high as N/16. I n experiments made by Connell, the bacilli grew well in urine of acidity below N/50; when the acidity lay between Nj50 and N/25, they were inhibited in varying degrees ;

REMARKS O N TYPHOID BACILLURIA. 367

and in the highly acid urines growth was slight. He observed that the degree of inhibition did not bear a constant relation to the degree of acidity, and experimented with urine originally of low acidity, artificially acidified by various substances. He was able to determine that a comparatively high acidity with acid sodium phosphate was required to inhibit growth, while acetic and lactic acids absolutely stopped growth when the acidity reached N/50, and even at N/500 exercised some inhibition. He concluded that the inhibitory factors normally present were the organic acids of unknown nature, which Folin (1 903 12) says are the cause of a varying amount, sometimes more than half, of urinary acidity.

It is supposed that bacilluria passes off by the washing of the bacilli out of the bladder by urine, after this has ceased to be a suitable medium for growth.

Bacilli other than B. typhosus have been found in the urine in enteric fever.

Blumer mentions eight cases in which bacilli were present. From the urine of six a pure culture of B. coli was obtained ; from the seventh, B. coZi and B. typhosus ; and from the eighth B. typhosus alone. H e does not mention whether or not these patients were females. Connell described the occurrence in typhoid urines of " colon bacilli, proteus, hay bacilli ('?), and mixed bacilli." One of these cases of colon bacilluria in a male was chronic. I n every other instance, he says, he was able t o satisfy himself that the colon, proteus, and mixed bacilli were contaminations, and that their presence was traceable to previous catheterisation. Jacobi (1902 13) obtained B. coli from the urine of several out of thirty cases of enteric fever, but does not say whether the patients were male or female.

My experience of the phenomena of typhoid bacilluria agrees in the main with those which have been quoted. Fifty-eight male patients were observed throughout the course of an attack of enteric fever, and bacilluria visible t o the naked eye was found to occur in seventeen cases. The bacilli were isolated from the urine on a modified Endo medium, with a reaction faintly alkaline to litmus instead of strongly alkaline (Kendall and Day, 1911 14), and were subjected to routine tests described later.

In every case bacilluria occurred as a casual phenomenon, and was unassociated with symptoms. A rise of temperature to 100" Fahr. took place in one patient on the first two evenings on which bacilli were present in the urine, but in no other case was there pyrexia or other disturbance attributable to this cause. The features of all the cases were the same. Without warning, and without apparent cause, bacilli suddenly appeared in the urine, and produced turbidity corre- sponding to, roughly, the presence of from 150,000,000 to 500,000,000 per C.C. This turbidity sometimes persisted for days, sometimes dis- appeared spontaneously or under treatment, and sometimes re- appeared as suddenly as it had originally come, but in no instance became chronic.

368 ADAM PA TRICR.

The bacilli appeared first in the urine a t the following times :-

TAULE I.- Time of First Appearance of Bacilluria.

Time. First week of pyrexia . . Second ,, ,, . . Third ,, ,> . . Fourth ,, 3 , . *

Fifth ,, I , . . Eighth ,, I , . . First week of convalescence . Second ,, I

Third ,, , Total .

I No. of Cases.

. 1

. 1 * 1

. 1 7

Time. Third week before apyrexia . Second ,, First ,, First week of coiiviilesceiicc . Second ,, , Third ,,

2 ,

Total .

No. of Cases.

. 2

. 3

. 6

. 2

. 3

. 1

. 1: -

The earliest case occurred on the tenth day of pyrexia, and the latest on the sixteenth day of convalescence.

Pus was present with the bacilli in t w o cases -only, and a little albumin in these two, and in two others. I n another case a little pus appeared for a day some time after the bacilluria was established. The diazo reaction was positive in only two instances,-the first and the third of the series.

With regard to the duration of bacilluria, some of these cases were notable for the quite short time during which it was present. I n three cases it lasted for one clay only, in three for two days, and in one for three days. I n these seven patients the condition passed off without treatment, and did not recur. I n another case i t disappeared or1 the second clay, and, recurring on the third day, persisted until urotro- pine was given on the tentmh day. In the other cases the urine cleared up after the administration of from 20 to 50 grs. of urotropiiie by the month, within thirty-six hours of the institution of the treatment, bacilluria having lasted for periods varying from two to fifteen days. It recurred in one patient for two days, in one for three days, and in a third for one day on three occasions, but in all these i t disappeared without any drug treatment.

The reaction of the urine was estimated in the case of M.B. from the thirtieth till the fiftieth day. The primary pyrexia here lasted for twenty-seven Clays, and the teniperature was elevated during a relapse from the thirty-fifth till the fifty-fifth day. Bacilluria appeared first on the twenty-fifth day. Thereafter it was present on three single days,--thirtieth, thirty-seventh, and forty-sixth. On these days the reaction of the urine was N/22, Nj25, and “33 ; while on other days, when the urine was less acid and presumably more suitable for growth, they did not appear.

Phenol-phthalein was t h e indicator used.

REMARKS O N TYPHOID B A CILL URIA. 369

It must be said that the explanations of the occurrence of bacilluria offered by the writers on the subject are unconvincing. Where abscesses have formed in the kidney it is reasonable to suppose that the bacilli come from these abscesses, but pus cells also would necessarily be present in the urine. In one of my patients who died, multiple small abscesses, from which B. typlzoszu was obtained in pure culture, were present in the left kidney, and a small amount of pus, which also gave a pure growth of B. typlzosus, had collected in a dilatation of the left ureter, just before its entrance into the bladder. Pus was present in the urine in considerable quantity. The ordinary cases of bacilluria, however, in which no pus is present in the urine, cannot be explained on the assumption that such a pathological con- dition exists in the kidney.

The evidence brought forward for infection of the bladder alone, without involvement of the kidney, is scanty, but so far as it goes, it shows that this niay sometimes occur.

If, as is supposed, bacilluria disappears by the washing of the bacilli out of the bladder, then the view cannot at the same time be held that the organisms enter the urine early in the disease, and remain there for days without multiplying, until the reaction of the urine becomes suitable for their growth. In such a case these organisms would be washed out in the same way.

However the infection occurs, it is certain that growth does not take place in the urine in the simple way described. A patient often passes every four hours for days, urine of such turbidity from the presence of bacilli, as can be obtained by growth in a fluid medium only after fifteen to eighteen hours' incubation. It is difficult to obtain a growth of more than 750,000,000 bacilli per C.C. in bouillon at 37" C. in thirty-six hours, and urine with 400,000,000 in each C.C. is often passed in cases of bacilluria.

A more feasible explanation is that the bacilli grow on the wall of the bladder, as they would on a solid culture medium, and that some are constantly washed off into the urine. This conception of their growth on the mucous membrane is compatible with their growing only in contact with urine of a suitable acidity.

The seventeen strains of bacilli were tested by the methods indicated below. Of these seventeen, eleven showed the following characteristics, and wore therefore regarded as typhoid bacilli :-

Production of acid, without gas, in glucose, maltose, and mannite. Production of slight permanent acidity, without clotting, in litmus

Non-fermentation of lactose and saccharose. Non-production of indol in peptona yater after seven days' growth. Non-liquefaction of gelatin. Colourless growth on potato. Presence of motility.

2 5 - ~ ~ . OF PATII.-VOL. XYIII.

milk.

3 70 ADAM PATRICK

B. typltosus . . . 1. W.B. . . . . 2.W.C. , . . . 3 . A . R . . . . . 4.R.S . . . . .

6. P.M. . . . . 5. C.D. . . . .

______--

The remaining six strains showed the reactions which are given below :-

TABLE IT.-Reaction of Atypical B a c Z i (February 1912).

(All gas formation slight.)

I A I A A

A 1 A AC

A ~ A A

A ( A A(

A ' A A

A ** A A( A( I

Litnius Milk.

d $ 2 ~n d n -- A A

A A

A A

A A

A A

A A

AC AC

-

$ a L n ?. -

A

Alk

Alk

Alk

A1 b

AC

AC

Time of growth-ten days.

+ = Presence.

A = Formation of acid. C=Formation of clot.

AG = Formation of acid and gas, Alk= Formation of alkali.

With the exception of strain C.D. they all formed an abundant brownish- yellow growth on potato. The bacillus C.D. grew on potato as a translucent streak, like B. typhosus.

On the modified Endo medium used they all formed in twenty-four hours colourless translucent colonies, 1 to 1.5 mm. in diameter. I n the case of C.D. and P.M., these colonies within thirty-six hours showed a red centre, and within forty-eight hours a red halo was appearing in the surrounding medium.

The six bacilli were considered to belong to two types-(1) A type not fermenting lactose and saccharose, forming alkali in milk, and showing abundant brownish-yellow growth on potato ; ( 2 ) a type fermenting lactose, forming acid a t least in saccharose, and acid and clot in milk.

The bacilli W.B., W.C., A.R., and R.S. were classified under the Jirst type, the differences among them being slight. W.B. produced a little gas as well as acid in mannite, and R.S. was of doubtful motility.

The bacilli C.D. and P.M. were attributed to the second type. C.D. produced a bubble of gas in saccharose as well as acid, and took several days t o clot milk, while P.M. brought about coagulation of milk within twenty-four hours. They differed as regards their growth on potato, but this is now usually considered to be an unreliable criterion.

Type 1 resembled most nearly the paracolon bacillus, which forms acid and gas in glucose and maltose, and does not ferment lactose and sacchnrose. The bacilli of Type 2 were thought to be allied to the B. co2i A group, which produces acid and gas in glucose, maltose, lactose, and saccharose. None of the six bacilli, however, was typical, for gas production was commonly absent, snd, where present, was slight in amount.

REMARKS O N TYPHOID BACILL URIA. 371

I t is to be observed that the bacilli of Type 1 showed the same reactions as B. typhosus in tests with glucose, maltose, lactose, saccharose, gelatin, and peptone water, while differing markedly in litmus milk and on potato. It is possible that such bacilli in enteric fever urines have previously been overlooked through the application of tests insufficient to ensure identification. The use of litmus milk as a test medium for bacilli from such urine seems advisable.

The cases in which these bacilli were found in the urine presented no points of difference from those of the ordinary type of typhoid bacilluria. In the earliest case (P.M.) the bacilli appeared on the twenty-first day, but the temperature had then been normal for five days. I n the others, bacilluria occurred respectively on the twenty-eighth day (first of normal temperature, W.B.), thirtieth day (last but one of pyrexia, R.S.), thirty-eighth day (sixth of normal temperature, A.R.), forty-ninth day (sixteenth of normal temperature, W.C.), and the fifty-seventh day (tenth of normal temperature, C.D.).

The bacilluria was commonly of short duration, lasting in three instances one day (W.B., A.R., and P.M.), in one instance two days (C.D.), in one three days (W.C.), and in the sixth instance (R.S.) for ten days, with a day's inter- mission on the second day, until it disappeared permanently after the admini- stration of urotropine.

As the facts tabulated below will show, there was no doubt that the patients from whose urine these six bacilli were isolated suffered from enteric fever, although cultures from the blood and fmes were negtztive, except in the ease of the blood of W.B.

1. W.B.-Pyrexia lasted twenty-eight days. Rose spots present. Pulse rate comparatively slow (temperature, 103O.6 ; pulse rate, 84). Diarrhea present. B. typhosus grown from the blood. Widal reaction positive (limit, 1-3000).

2. W.C.-Pyrexia lasted thirty-three days. Rose spots present. Pulse rate comparatively slow (temperature, 101O.6 ; pulse rate, 84). Spleen palpable. Widal reaction positive (limit, 1-500).

3. A.R.-Pyrexia lasted thirty-two days. Rose spots present. Pulse rate comparatively slow (temperature, 103"*4 ; pulse rate, 96). Diarrhea present. Widal reaction positive (limit, 1-800).

4. R.S.-Pyrexia lasted thirty-three days. Rose spots present. Pulse rate comparatively slow (temperature, 103O.6 ; pulse rate, 92). Diarrhea present.

Pulse rate comparatively slow (temperature, 103".2 ; pulse rate, 84). Diarrhea present, with pea-soupy motions. Spleen enlarged. Widal reaction positive.

6. P.M.-Pyrexia lasted sixteen days. Rose spots present. Pulse rate comparatively slow (temperature, 103".8 ; pulse rate, 94). Widal reaction positive (limit, 11-200).

Five of the bacilli (the exception being C.D.) were tested with a stock anti-typhoid serum, obtained from a rabbit (limit of agglutination, 1-80,000), and with the serum of the patient from whom each was isolated. As the table shows, four of the five were unaffected by the anti-typhoid serum, while they were agglutinated in varying degrees by the Serum of the patient,-in one instance (W.B.) to a greater extent than was the laboratory strain of B. typhoszcs by the same serum.

The agglutination of these bacilli from the urine with the patient's serum, and not with the anti-typhoid serum, would seem to point either t o some relationship between the bacillus from the urine and

Widal reaction positive (limit, 1-1800). 5. C.D.-Pyrexia lasted forty-seven days.

372 ADAM P A TRICK.

Limit with Patient's Serum.

the typhoid bacillus which caused the fever ; or to a double immunisa- tion, both with typhoid and this bacillus. The bacilli, however, were isolated on the first day of their appearance in the urine, and in the two instances in which the serum reaction was most active the test was carried out on the second day after the bacillus first appeared. As agglutinins are not produced to any extent until the sixth day, it is evident that if a double inimunisation took place, the bacilli which appeared in the urine must, for a t least some days previously, have been exercising an immunising influence on the patient.

Limit of Agglutination of Stock Typhoid

Bacillus by

TABLE HI.-Agglutination of the Atypical Bacilli.

32ncl 37th

of normal omperature.

34th

Strains.

~

1. W.B.

2 . R.S. 3. P.M.

4. W.C.

5 . A.R.

2 32

18

1 Devs after Day of I Isoiation of

Bacillus when

1 Tested. Test.

Limit with Anti-tgphoid

Serum.

~

No agglutination at 1-25

1-50

No asrsrlutination 1-70 1 1-500

The fermentation and other tests to which the atypical bacilli were subjected were carried out between November 1911 and February 1912. They were kept in an ice-cupboard for a year on agar slopes at a temperature of from 5" to 7" C., and were subcultured a t most twice. They were exaniined again subsequently to February 1913, when the range of tests was extended, and the organisms grown i n various other fermentable substances. T h e final results are shown i n Table IV., together with the results of the tests with B. t y p h u s in the same media.

The bacilli showed certain features in coinnion. They were all Gram-negative,-all formed acid in glucose, and acid and gas in maltose, niannite, saccharose, gdactose, lr-evalose, rhamnose, glycerin, and inulin. B. typhosus does not ferment saccharose, rhamnose, or glycerin, and produces no gas in any of the media employed. All six failed to ferment erythrite, adonite, dulcite, and dextrin ; none liquefied gelatin, and Vosges and Proskauer's reaction was negative in each case.

Strain P.M. was distinctly in a class by itself. It aloiie fermented lactose, raffinose, and amygdalin, forming acid and gas in the two former, and acid in the last; and it was the only bacillus to produce indol, and to form acid and clot i n litmus milk.

TA

BL

E IV

.-Rea

ctio

ns

of At

gpic

al B

acil

li (

1913

).

Litm

us M

ilk.

I I St

rain

s.

I I

$3

0 4 g p: -

A

AG r: AG

AG

AG

AG

a5 U

F

Ld .- z ._

A

AG

AG

AG

AG

AG

AG

i s J)

'=

-

A

AG

AG

AG

AG

AG

AG

I I I ~~ I I B. t

?J>

hosz

ts

I

1.

W.B

. .

2.

W.C

. .

I I 3. A.R. .

~ 4.

R

.S.

.

I 5.

C.D

. .

6. P

.M.

.

-~

4 n ri

-

A

A

A

A

A

A

AC

-

A

AG

AG

AG

A

G

AG

A

G

AG

A

G

AG

AG

AG

AC

;

+ t- ~

+' I

AG

AG

AG

~

AG

I

AG

/ AG

AG

A

G

AG

AG

AG

AG

-i

-

i -1

- +

j-

- I

fl

+I

-! A

=For

mat

ion

of a

cid.

C =

Fur

mat

ioil

of c

lot.

AG

= F

orm

atio

n of

aci

d an

d ga

s.

Alk

= F

orm

atio

n of

alk

ali.

+ =P

rese

tice.

374 ADAM PATRICK.

The five other bacilli showed that they were closely allied to one another, and four (W.U., W.C., A.R., and C.D.) give identical results in all the tests used. Perhaps the most charactcristic feature of the f ive was t h e production of alkali i n litmus milk after a few days’ growth. This sharply distinguished them from the typhoid group, which produces permanent acidity, and from l’.AI., which produced acid and clot.

As has been said, W.B., W.C, A.R., and C.D. ultimately gave the same results. I n tlie first culture i n inulin, however, C.D. formed acid without gym, but in a subcultnre from this produced also gas. W.C. a t first formed acid without gcns i n gnlactose antl lmwlose, but when a fresh culture was made from the agar tube, a month later, 91s also was produced. Similarly, a t first, A.K. produced 110 gas i n saccharose, but in a fresh culture after two months produced gas as well as acid. It was non- inotile, :uid did not ferment arabinose, in which all the others produced acid and gas.

B.8. showed a greater divergence.

A subsequent culture a month later yielded the same result.

Tlie result of the tests in 1913 yielded iniportant differences froiu those carried out i i year earlier, and showed thaLt the bacilli were in a somewlut unstable condition as regards fermentative powers.

The oliange in tlie cliaracters of the bacillus C.D. is difficxlt of explanation, except on the hypothesis that originally i t included two strains,-one resembliug P.X, the other, the four other bacilli,-and that^ tlie former strain died out. Origindly (3.1). fermented la.ctose, formed indol, and produced acid and clot in milk. il year lator i t did not ferment lactose, did not produce indol, and formed alkali in milk.

Tlie changes in the other bacilli were all in the direction of greater fermentative puwrs , though they hati been cultivated simply on ordinary agar. Whercm a t first they all formed acid only in maltose, a year later gas as well as acitl repulted from their growth. All now also produced gas as well as acid in iiiilnnite, whereas formerly W.C. and R.S. produced only acid. A marlied change took place with regard t o their action on saccharose.

In the original tests in saccharose (1912), strain P.M. produced acid, C.D. acid with n bubble of p s , while the other bacilli left the sugar unchanged. When tho tests were repeated in 1913, strain P.M. produced gas as well as acid, aiid C.D. and ILS. produced acid. Gas production with P.M. and acid production with R. S. were new characters, and therefore successive subcultures were made with all the bacilli t o see whether further action on saccharose resulted. A second series of subcultures was made a t the end of ten days, and a third from the second a t the end of a further ten days.

W.B. and W.C. in the first subculture were unchanged ; in the second they formed acid, and in the third acid aiid gas.

(2.1). and R.S. in the first subculture formed acid, and in the second acid and gas.

A.R. in the first subculture was unchanged, and i n the second, third, and fourth formed acid without gas.

The Bacillus A.R. was kept for two months longer on agar a t 5’ to 6” C., antl a t the end of that time was found to produce gas as well as acid in t h e first subculture.

It was found also that acid tended to be produced more rapidly, and gas in greater volume, by all the organisms, with successive subcultures. For instance, R.S. in the first subculture produced acid at the end of ten days; i n tlie secoiid, acid and some gas a t the end of four days ; and in the third, acid after two days, and subsequently a larger volume of gas.

These results are shown in the subjoined table :-

Strains.

1. W.B. . 2. W.C. . 3. A.R. . 4. R.S. . 5. C.D. ,

6. P.M. .

B. typhosus . . 1. W.B. . . 2. W.C. . . 3. A.R. . . 4. R.S. . .

REMARKS ON TYPHOID B A CILL URIA.

TABLE V.--Reaction of Atypical Bacilli in Saccharose.

- A (slight) A (slight) A (slight) A

February 1912. First

Subculture.

(Ten days' growth a t 37" C.)

First hibculture.

-

A

-

A

A

AG

A = Formation of acid. AG = Formation of acid and gas.

Febrnary 1913.

Second Subculture.

A

AG

A

AG

AG

AGt

Third lubcultnre.

Fourth hbcultiire.

...

... A

...

...

...

375

April 1913. First

Subculture.

...

... AG

...

...

...

*I- Gas formation very slight. 1. =More gas than in previous subculture.

In the tests with glycerin (1 per cent. in peptone water) an increase of fermenting power developed in the second subculture. A t the end of five days no change was visible in the colour of the litmus with any of the bacilli. After ten days, however, all except P.M. showed a tendency to form acid, the litmus having become of a purple tint. In the next subculture (made from the first) all except P.M. produced slight acid and a little gas. P.M. showed no tendency to bring about fermentation. In the third subculture, however, it produced slight acid and a little gas a t the end of ten days.

The results of the tests with glycerin are shown in Table VI.

TABLE VI.-Atypical Bacilli in Glycerin. (Ten days' growth a t 37" C.) ' Strains. First Subculture. Second Subculture. i

A (slight) 6. P.M. - 5. C.D. .

- AG AG AG AG AG -

I 1 A =Formation of acid. AG=Formation of acid and gas.

1 Third Subculture.

I - I

1 AG AG AG AG AG AG 1

As has been mentioned, the bacilli were considered at first to belong to two types-( 1) one resembling B. paracoli, (2) one resembling B. coli A.

376 ADAM PATXICK;

The power which the bacillus P.M. developed of forming gas as well as acid in saccharose brought it into conformity with the latter type, which, according to Wulff (1 9 1 1 15), produces acid and gas in lactose, glucose, maltose, and saccharose, and also in galactose, xylose, and niannite. As regards the other bacilli, their development of the power to ferment saccharose made their classification as paracolon bacilli doubtful, as B. paracoli is not a saccharose fermenter. If they were paracolon bacilli, then they were developing new characteristics. The possibility that they were modified typhoid bacilli is unlikely, but in any case their sudden appearance in large numbers in typhoid urines was a curious phenomenon. If they were not altered typhoid bacilli, then it is practically certain that their origin was the intestine, and the facat that they were agglutinated by the patient's serum, though not by the anti-typhoid serum, seemed to point to an immunising action of the bacilli themselves, or their toxins, on the body.

The occurrence of mutations in micro-organisms has been widely investigated since Massini's (1 9 0 6 16) account of his B. coli nmtabile in 1906. Apart from such sudden development of new characteristics, it is well known that certain properties can be produced as a result of training. Twort (1 9 0 7 17) showed that some typhoid-coli organisms acquired the power of fermenting certain sugars by long growth in them. For instance, he trained 3. typhosus to ferment lactose and dulcite.

The changes which I have described in these atypical bacilli were of two kinds: (1) those which took place as a result of training, and (2) those which occurred spontaneously. Of the former kind was the power of A.R. to form acid, and of W.B. and W.C. acid and gas in saccharose, of C.D. to form gas in inulin, and of P.M. to form acid and gas in glycerin. But it is to be noted that the production, by all, of gas in maltose, by W.C. of gas in mannite, by P.M. of gas in saccharose, and by R.S. of acid in saccharose were new characteristics which developed during the year the bacilli were stored on agar slopes at a temperature of 5" to '7' C., and without any contact with the sugars. The bacillus A.R., also, which failed to produce gas in saccharose in a series of four subcultures made at intervals of ten days, developed this power when kept for two months longer on agar a t a tempera- ture of about 6"C., forming gas as well as acid in the first sub- culture. , Similarly, W.C. acquired the power of forming gas in addition to acid in galactose and levdose when kept for another month a t 6°C.

Typhoid bacilli freshly isolated.froni the body are frequently non- agglutinable, and instances are recorded (Nicolle and Trenel, 1 9 0 2 l8 ; Lipschul,z, 1904 lo) of the development of agglutinability by non- agglutinable bacilli kept without subculture for three months ; but change in fermentative properties has commonly been described as occurring after a course of training. Sorensen (1 9 1 2 g o ) , however, has described a case of glycosuria in which an organism, B. pwumaturim,

REMARKS O N TYPHOID BACILLURIA. 377

was isolated from the urine. This bacillus produced gas in the bladder, and also in artificial media containing glucose, lactose, and saccharose. After two years, gas ceased to be produced in the bladder, and it was found that the organism had lost the power of forming gas in sugar-containing media. A year later the bacillus suddenly reacquired the power of forming gas, and shortly afterwards the patient began to suffer again from pneumaturia. These changes took place spontaneously.

While it is unlikely that the atypical bacilli were altered typhoid bacilli, a mutation form of B. typhosus has been described by Mandelbaum (1 9 0 7 "'>. This organism, which he called B. metatyphi, was isolated from the fzces of a typhoid carrier. It differed from B. typhosus chiefly in forming acid in the presence of glycerin, instead of alkali, and was considered by Mandelbaum to be a mutation form of B. typhosus produced by growth in the body. On glycerin-agar plates some colonies of B. metatyphi produced papillz of B. typhosus.

iVote o n Bacilluria in Female Patients with Enteric Fever. The seventeen cases of bacilluria which have been described

occurred in men. I had afterwards an opportunity of examining the urine of a few women suffering from enteric fever.

Bacilli were found to be present in large numbers in the urine of seven women, the physical characters of the urine being such as have been described for typhoid bacilluria. I n five cases bacilli were present in convalescence on the day on which the urine was first examined. In the sixth case bacilluria was known to exist a t the onset of the fever, and in the seventh case the bacilli appeared first in the urine on the twenty-seventh day. I n this last case the bacilli were typhoid bacilli, in the six others B. coli.

SUMMARY OF RESULTS.

1. It was found during the investigation of bacilluria in men suffering from enteric fever, that in six cases out of seventeen the bacilli were not typhoid bacilli, but members of the typhoid-coli group not previously described. It has always been supposed that the organisms present in " typhoid " bacilluria are typhoid bacilli, or some contaminating organism, such as B. coli. B. coli did not occur in any of the seventeen cases. In seven instances of bacilluria in women with enteric fever, on the other hand, the bacillus present was found, in all but one, to be B. coli.

2. These atypical bacilli appeared in the urine, on the average, rather after the date for typhoid bacilluria, and persisted for quite a short time,-in three instances for one day only. Their appearance was not attended by any constitutional disturbance, and in general the cases resembled ordinary cases of typhoid bacilluria.

378 REMARKS O N TYPHOID BACILL URIA.

3. These bacilli, though practically unaffected by artificial anti- typhoid serum, were agglutinated in varying degrees by the serum of the patient from whom each was isolated,'in one instance to a greater extent than was the stock typhoid bacillus by the same serum.

4. The six bacilli were stored on agar slopes a t a temperature of about 6"C., and on re-examination a year later were found to have acquired greater fermentative powers with regard to certain sugars. This property has usually been described as developing by continued growth and subculture of an organism in a solution of the sugar, but with these bacilli the power, though afterwards increased by sub- culture, developed in many instances spontaneously.

REFERENCES,

1. PETRUSCHKY , . . . .

2. GWYN . . . . . . . 3. CONNELL . . . . . . 4. BUCRAN . . . . . 5. LIEBETRAU . . . .

6. HORTON SMITH. . . . 7. KONJAJEFF . . . . .

8. FLEXNER . . . . . . 9. BROWNLEE AND CHAPMAN

10. BLUMER . . . . . . 11. P A R K . . . . . . .

12. FOLIN . . . . . . . 13. JACOBI . . . . . . .

14. KENDALL AND DAY . .

16. MASSINI . . . . . .

17. TWORT. . . . . . . 18. NICOLLE A N D TRENEL. ,

19. LIPSCHUTZ . . ,

20. SORENSEN . . . . . . 21. MANDELBAUM . . . .

Centralbl. f. Bakteriol. u. Parasitenk., Jena, 1898, Abth. I., Orig. Bd. xxiii. S. 577.

Johns Hopkins Hosp. Bull., Baltimore, 1899,

Am. Journ. Med. Sc., Phila., 1909, vol.

LiverpoolMed.Chir. Journ.,1908, Jan., p. 166. Arb. a. d. kaiser. Gesundheitsniiate, 1906, Bd.

Lancet, London, 1900, vol. i. p. 910. Abstract in Centralbl. f. Bakteriol. u. Para-

sitenk, Jena, 1889, Bd. vi. s. 672. Journ. Path. and Bacteriol., Cambridge, 1896,

vol. iii. p. 202. Glasgozo Med. Journ., 1906, vol. 1 x 6 p. 407. Johns Hopkins Hosp. Rep., Baltimore, 1895,

Trans. Assoc. Am. Phys., 1901, vol. xvi. p.

Am. Journ. Phys., 1903, vol. ix. p. 265. Deutsclzes Arch. f. klin. Med., Leipzig, 1902,

Journ. Med. Research, Boston, 1911, vol.

Centmlbl. f. Bakteviol. u. Parasitenk., Jena,

vol. x. p. 109.

cxxxvii. p. 637.

xxiv. S. 341.

vol. v. p. 327.

634.

Bd. Ixxii. S. 442.

xxv. p. 95.

1911, Abth. I.. Orig. Bd. lxv. S. 27. Arch. f: Hyg., Munch& u. Leipzig, 1906, Bd.

Ixi. S. 250. Proc. Ro?y. Soc., London, 1907, vol. lxxix.

B., p. 329. Ann. de I'lnst. Pasteur, Paris, 1902, tome

xvi. p. 562. Centralbl. f. Bakteriol. u. Parasitenk., Jena,

1904, Abth. I., Orig. Bd. xxxv. S. 798. Ibid., 1912, Abth. I., Orig. Bd. lxii. S. 582. Munchen. iiied. Wch,nschi*., 1907, Bd. liv.

S. 1766.