No. 3905

JULY 2, 1898.THE LANCET,

The Croonian Lectures ON

THE CHEMICAL PRODUCTS OF PATHOGENICBACTERIA CONSIDERED WITH SPECIAL

REFERENCE TO ENTERIC FEVER.Delivered before the Royal College of Physicians of London

on June 14th, 16th, 21st, and 23rd, 1898,

BY SIDNEY MARTIN, M.D.LOND., F.R.S.,FELLOW OF THE COLLEGE; PROFESSOR OF PATHOLOGY, UNIVERSITY

COLLEGE ; AND ASSISTANT PHYSICIAN TO UNIVERSITY COLLEGEHOSPITAL AND TO THE HOSPITAL FOR CONSUMPTION,

BROMPTON.

LECTURE IIL1

Delivered on June fZ 1st.

PRODUCTION OF THE POISONS OF THE TYPHOID BACILLUSIN ARTIFICIAL-CULTURE MEDIA.

MR. PRESIDENT AND GENTLEMEN,—Two different kinds

of media were employed in which to grow the bacillus.

Ordinary peptone broth was used to determine the presenceor not of secretory products of the bacillus and broth con-taining no peptone but mixed with some proteid solution-either blood serum or alkali albumin made from the

spleen or extract of lymphatic gland-in order to deter-

mine the presence of digested products of the bacillus. In

some cases liquid blood serum sterilised at 60° C. was used

without the addition of broth. The virulent bacillus growswell in ordinary peptone broth but the growth in solutionscontaining much proteid is very much less exuberant. Thisis especially the case with sterile blood serum and most

commonly the bacilli, after growing for a time, sink to thebottom in clumps and cease growing, although they still arealive and active. They grow much better in a solution

oontaining a small proportion of alkali albumin, butstill not nearly as well as they do in broth. The objectof growing them in the presence of proteids was totest whether there was any digestive action of thebacillus. In one experiment two test tubes were filledabout 3 in. deep with liquid blood serum, which was

rendered semi-solid by heating at a temperature of from 65°to 70° C. The medium was translucent and one of the tubeswas inoculated with the bacillus, the other being kept as acontrol. It was observed that the bacillus rapidly grew inthe medium, spreading through all parts of it, but that noliquefaction indicative of digestion occurred during incuba-tion, which lasted 102 days. When grown in a liquidmedium, however, containing a small proportion of alkalialbumin, and especially alkali albumin made from splenicpulp, the virulent bacillus has a slight but very insignificantdigestive action. Thus in one experiment after thebacillus had grown 32 days in such a medium it was foundthat the solution gave a distinct biuret reaction with coppersulphate and potash and that the alkali albumin was

beginning to be changed into albumoses. This digestiveaction, although slight, is explanatory of the albumosesfound in the spleen in cases of typhoid fever.

Poison formed in proteid solrtiorzs.-If the non-virulentbacillus is grown in a solution containing digestible proteidinstead of peptone a poison is formed which has an actionlike that of the poison formed in broth. Thus in one

experiment the bacillus was grown in a 1 per cent. solutionof serum alkali albumin in broth for 29 days, at theend of which time it was filtered, the filtrate con-

ce’ttrated, and the proteid was separated by alcohol. Someamount of digestion had taken place as the solution gavea biu-ret reaction. The solid product obtained was injectedin two doses (about 0’3 gramme) on successive days into themarginal vein of the ear of a rabbit. There was a marked

lowering of the temperature and a subsequent rise followingeach injection, but not to the average normal temperatureof the animal for some days. Some bodily depressionwas observed on the first two days of the experiment and

1 Lectures I. and II. were published in THE LANCET of June 18thand 25th, 1898, respectively.

No. 3905

there was a loss of body weight, but there was no diarrhoeaand the animal subsequently completely recovered. Inanother experiment the non-virulent bacillus was grown in asolution of alkali albumin made from the splenic pulp butwithout adding broth. One flask was filtered after 45days and 10 c.c. of the filtrate were injected in two doses onsuccessive days into the marginal vein. After the first

injection there was only a very slight rise of temperaturewhich, however, was more marked after the second injection.The animal showed no particular symptoms except a gradualloss of weight, which amounted to 235 grammes at death, onthe twenty-eighth day of the experiment. The post-mortemexamination revealed no naked eye change in any of theorgans or tissues of the body. No cultivations were obtainedfrom the heart’s blood ; the heart muscle showed no fattydegeneration and there was no degeneration of the peri-pheral nerves. Another culture was filtered after 59 days’incubation, and 12 c.c. were given in two doses on successivedays also into the marginal vein. There was a well-markedrise of temperature after each injection, but no diarrhoea wasproduced and there was only a slight loss of weight whichwas recovered from. The animal died in 24 days and nocultivations were obtained from the heart’s blood, liver,spleen, or peritoneal cavity. There was subcutaneous hoemor-

rhage in patches in front of the thighs and in the caudalregion, and the peritoneum contained about 1 centimetre ofclear fluid. There was no peritonitis and the organs showedno change to the naked eye. The results of growing the non-virulent bacillus in a proteid solution show that the poisonsare of the same nature as that formed by the virulentbacillus in broth. The digestion which takes place in theproteid solution is extremely slight and the toxic body isthus mainly a secretion of the micro-organism.Although the toxic products are developed in the presence

of digestible proteids, yet they are most evident when thebacillus is grown in peptone broth. In one experiment thebroth was inoculated with the typhoid bacillus obtainedfrom the spleen in a case of enteric fever but before it hadbeen made virulent by passage through animals. The

growth was allowed to continue for 11 days, and thebroth was filtered through a Chamberland filter. Two injec-tions into the marginal vein were made on successive days.The first injection, of about 3 c.c., resulted in the loweringof the temperature and in the production of diarrhoea soonafter the injection. The second injection-also of 3 c.c.-produced but little effect, and on the three subsequent daysno appreciable change occurred in the animal. On the six-teenth day of the experiment the animal was found to havelost 170 grammes in weight, and to have some degree offever, which, however, did not reappear. It lived for 24days gradually losing weight until at death it had lost235 grammes. At the post-mortem examination therewas no obvious lesion found in any of the organs of thethorax or abdomen, the intestines were normal and culturesmade from the heart’s blood in the peritoneal cavity weresterile, showing that there was no bacterial infection. Theheart-muscle, examined after staining with osmic acid,showed degeneration of most of the fibres, there being loss ofstriation with longitudinal fibrillation and fatty granules inthe muscle cell. This experiment shows several points whichwill be brought out more fully later: (1) that a poisonoussubstance is present in the broth apart from the bacillusitself ; (2) that this body produces a lowering of temperatureand diarrhoea, as well as a loss of weight; and (3) thatat death there was distinct degeneration of the cardiacmuscle. ’

The broth culture from the same stock of bacillus was more

poisonous and produced a more profound effect if the incuba-tion lasted 25 days. A rabbit received a single dose of

--

6 c.c. of the broth filtrate after 25 days’ growth. The initial

, effect of this injection into the venous system was a rise of

.

temperature (the maximum being attained in two and a half, hours) and a rapid fall, during which there was profuse: diarrhoea, with collapse and laboured breathing. This,

period of depression lasted some hours and was succeeded: on the next morning by a febrile rise which continued

l during the whole of the day. On the following day the. temperature was somewhat irregular, but subsequently no, great change in the temperature was noticed. The weight.

fell rapidly up to the fourth day of the experiment, when theanimal had lost 120 grammes. Subsequently, however, it

regained part of the weight, and at death, on the twenty-i sixth day of experiment, the loss was only 35 grammes from

the original weight. No naked-eye changes were observed onA

2

post-mortem examination and no cultures were obtained from may be that the heat breaks up the bodies and liberates thethe organs of the body. The heart-muscle showed slight poison inside, and after heating some of the bodies of thesigns of degeneration. bacillus may be observed disintegrated under the microscope.

The more virulent the bacillus the more toxic becomes the There is no evidence from the experiments that there are twobroth. Thus, after growing the virulent bacillus for 11 days kinds of poison, one extra-cellular or secreted by the bacillusthe micro-organism was killed by chloroform and the broth and one intra-cellular, with different actions. The toxicmixture centrifugalised, the clear broth being used for the properties of the broth from which the bacilli are separatedinjection. Five cubic centimetres were injected in two doses by means of the Chamberland filter are practically the same,of 2 c.c. and 3 c.c. respectively on successive days. An although not so pronounced as those observed when the brothinitial depression of temperature followed by fever was contains the bodies of the bacillus as well. This is notobserved after both injections, and the animal died in 12 days an unimportant point, inasmuch as some have made a dis-with no naked-eye change in the organs of the body, which tinction between the intra-cellular and the extra-cellulargave no bacterial growth. There was a very well-marked poisons of pathogenic micro-organisms. No effect is moredegeneration of most of the muscle fibres of the heart. marked or more constant in the action of the poison thanThese results may be taken to illustrate the action of the the production of diarrhoea which with the more activeclear broth without the bodies of the bacillus-i.e., of the poison is very profuse, lasting a varying time, sometimeschemical poison which is present in solution in the broth. hours, sometimes a day, and consists in the passage of

Action of broth with the bodies of the bacillus.-The toxic liquid motions, with mucns occasionally, but no blood ataction of the bodies of the bacillus itself was tested by any time. On a post-mortem examination of an animal

killing the bacillus with chloroform. When the culture was dying from this diarrhoea characteristic appearances are

sterile the chloroform was removed in vacuo and the liquid almost invariably found. The stomach may be full ofused for injection. In one experiment a 13 days’ old culture undigested food, the csecum and colon contain liquid fsecalwas used and 4 c.c. were injected into the marginal vein of matter, but the greatest change is in the small intestinethe ear. There was a fall of temperature in two and a half which is full of a slightly turbid, sticky fluid, containinghours, and this low temperature continued the whole of the but little food and no bile. This liquid is simply a solutionfirst day of experiment and also on the second day. On the of mucin, which may be separated by precipitation withmorning of the third day the mercury of the thermometer acetic acid. The mucous membrane is a little soft on thewould not rise and the animal died. Soon after the injec- surface and there is a very large increase in the number oftion there was slight diarrhoea. During the subnormal goblet cells. The Peyer’s patches are not apparentlytemperature the animal became increasingly weaker, the altered. The changes described by Sanarelli in the Peyer’sfur becoming ruffled, although there was no diarrhoea. The patches of the intestines have not been observed by me.loss of weight in the two days of experiment was 180 The loss of weight is also a constant symptom, and thegrammes. There was no naked-eye change in any of the more virulent the poison the more marked it is seen. It is

organs and the blood was sterile. This experiment may be also in these cases progressive. With a smaller dose of thetaken as showing the profound effects following the injection poison, however, the initial loss of weight is regained. Noof the poison. The great depression of temperature, how- naked-eye change is produced by the poison in the organs ofever, observed in this experiment is not the only effect to be the body. The effect of the poison, when death of theproduced. In some of the experiments with a weaker poison animal is slow, in producing degeneration of the cell fibre ofno depression ensued, but a well-marked rise of temperature ; the heart is well marked. The heart was never found in athus 3 c.c. of a broth culture (seven days’ growth) killed state of advanced fatty degeneration, but rather in thewith chloroform were injected into the marginal vein of the earlier stage, in which there is loss of transverse striation ofear of a rabbit. There was a considerable rise of the fibre, with the appearances of longitudinal fibrillationtemperature after the injection, which continued during and the presence of numerous granules in the fibre, staining

,

the whole of the following day and the day after. On the more or less black with osmic acid.fourth day the temperature partly regained the normal limit ETrEmrms ,..and continued afterwards normal. There was a loss of BACILLUS E’rERITlDIS (GARTNER).weight up to the seventh day after which the animal The experiments with this micro-organism were conductedrecovered. In this experiment no diarrhoea was produced in the same manner as those with the typhoid bacillus. Aand no symptom except that of fever and loss of weight. culture was kindly sent to me by Dr. Durham which gaveIn two or three other experiments with the broth and the the reactions previously described as characteristic of thebodies of the bacillus these features were noted, but there was bacillus. It differs from the typhoid bacillus in the factnot infrequently an initial fall of temperature, succeeded by that it causes gas formation in glucose-agar or glucose-a prolonged rise, and diarrhoea for a longer or a shorter period gelatin, but like the typhoid bacillus it does not coagulatewas a fairly constant symptom. There is no doubt, then, milk and it forms but little or no indol in broth. The originalthat the retention of the dead bodies of the bacillus in the culture of the bacillus worked with was toxic, as 5 c.c. of abroth renders it more poisonous. The effect on different two days’ old broth culture injected into the peritoneumanimals varies somewhat. In the experiment which was of a guinea-pig killed the animal in less than twenty-mentioned above 4 c.c. took two and a half days to kill. In four hours. The virulence of this bacillus was increasedanother experiment 4 c.c. of the same stock caused death by passage through animals in a manner similar to thatwith lowering of the temperature in three and a half hours. which has been described with regard to the typhoidThere is also some individual peculiarity in the animals as bacillus-that is, some of the peritoneal exudation fromregards their resistance against the poison. the first animal which died being injected into the

The effect of heat.-The effect of high temperature on this peritoneal cavity of a second guinea-pig, and from the

poisonous liquid containing the bodies of the bacillus was second into a third and so on. The death of eachtested in order to see whether the toxic substances were animal occurred as a rule in less than twenty-four hours,sensitive to heat or not. In one experiment the broth The dose was gradually lessened until a small loopfulwith the bodies of the bacillus was heated for ten of an agar culture injected into the peritoneal cavity killedminutes to 60° and 64C., 4 c.c. were injected into the the animal. In the series of experiments the intra-

marginal vein, and death occurred, with lowering of tempera- peritoneal injections caused the exudation of a highlyture and diarrhoea, in two hours and twenty minutes. This albuminous fluid containing numbers of bacilli. In thedeath was more rapid than in the control experiment with the earlier cases there was also an exudation of leucocytes ;same amount of broth unheated. As marked was the effect later on, however, these disappeared. Occasionally thereof exposing the broth to a higher temperature before injec- were a few flakes of lymph on the surface of the peritoneum,tion. 4 c.c. of a broth mixture heated at the boiling but only at the early stage of the series of inoculations. At

point of. water for five minutes were injected into the no time was there any exudation of blood. During themarginal vein and death with profuse mucoid diarrhcea series cultivations were made from the peritoneal fluidoccurred in two and three-quarter hours. This experi- of each animal to test the purity of the growth, to see

ment may be compared with a similar one previously whether the bacillus kept its characteristics, and also tomentioned in which the same amount of broth, but study the distribuon of the micro-organism in the body.unheated, was injected into another animal which lived It was seen that te growth was always typical and pure..for two and a half days with great depression of tem- and that when the bacillus became virulent it was foundperature. It may be said, therefore, that this degree of always distributed throughout the organs of the body, beingheat not only does not destroy the poison which is in the obtained from the heart’s blood, the spleen, the liver, and the.bodies of the bacillus, but actually brings out its effects. It kidneys and peritoneal cavity in pure cultivation. In this

3

respect the presence of the bacillus of Gartner in the various s sorgans of the body is more constant than that of the typhoid f fbacillus. t tThe peritoneal fluid was highly toxic. Thus in one t

experiment in which the bacilli were killed by means of tchloroform and removed by subsequent centrifugalising, 3 c.c. c

of clear peritoneal fluid injected into the marginal vein of a crabbit weighing 470 grammes caused death in one and a half r

hours, the temperature rapidly falling in half an hour from s

102° to 974°. The peritoneal fluid filtered through porcelain 1is not so toxic as the unfiltered fluid from which the bacilli 1have been removed by centrifugalising. This filtered peri- : atoneal fluid was found to produce fever when injected into 1the marginal vein of the ear of the rabbit, but when given in itwo doses of 1-5 c.c. and 1 c.c. it did not cause the death of Ithe animal nor any particular symptoms except a rise of

body temperature.Gro7vth of the bacillus in broth.-The experiments per-

formed in this way were precisely similar to those with thetyphoid bacillus. The virulent Gartner bacillus was grownin broth for 22 days and the culture was not filtered,but the bacillus was killed by chloroform. Experimentswere performed with the clear broth after centrifugalisingand with the turbid broth-that is, with the broth plus thebodies of the bacilli. Each animal received the same

dose-viz., a single dose of 4 c.c. into the marginal vein.The result of the injection of the clear broth was a con-tinued and marked depression of temperature during theday of experiment, the temperature next day again risingto normal. There was diarrhoea soon after the injectionof the poison and there was a progressive loss of weight. Asimilar dose of the broth, plus the bodies of the bacilli,caused a more marked depression of temperature whichlasted till the following morning and most of the followingday. Here also there was diarrhoea and a progressiveloss of weight. The effect of heat on the broth con-

taining the bodies is seen in the two following experiments.Thus the same dose (4 cc. of the mixture) heated to600 C. for ten minutes caused a depression of temperaturewhich was very great and ended in death in four and ahalf hours. This animal also showed diarrhoea. A similardose heated to 1000 C. for ten minutes produced also a

very great depression of temperature which lasted to someextent till the next day. There was well-marked diarrhoeaand great loss of weight. It is thus seen that thecharacteristic action of the toxic products of the Gartnerbacillus is very like those of the typhoid bacillus. There is

present in the broth from which the bacilli have beenremoved a poison which causes a great depression of

temperature, and this poison is also present in the bodiesof the bacillus and has an action similar to that of the

poison obtained from the typhoid bacillus—viz., it pro-duces diarrhcea and loss of weight. Like the poison ofthe typhoid bacillus, the virulence of a broth culture

containing the bodies of the bacillus is increased by heatingthe solution at 600 C., so that, as seen in the experimentsjust quoted, whereas 4 c.c. of the unheated mixture did noteause death, when it was heated to 60° for ten minutesdeath occurred in 32 hours with great lowering of the bodytemperature. The mixture heated to 100° C. for the sametime gives a greater effect than the non-heated mixture, butdoes not produce death in the same day as when the mixtureis heated to only 60° C. The effect of heat appears to be tobreak up the bodies of the bacillus so that the poison is

liberated, making the liquid more toxic. When grown in aproteid solution containing alkali albumin or serum proteidsGartner’s bacillus causes some amount of digestion with theformation of albumoses. This is more marked than is thecase with the typhoid bacillus, although the digestion isstill far behind that which occurs in the case of the bacillusdiphtherias and the anthrax bacillus. One result obtained isinteresting inasmuch as the bacillus coli also produces it-viz., that when grown in Marmorek’s fluid Gartner’s bacillusproduces a partial precipitation of the proteid solution in theform of a gelatinous clot. In strong solutions containingserum this result, however, was not observed and it was notso constant as with the bacillus coli.

BACILLUS COLI COMMUNIS.

There is one important difference between this bacillus.

and the two other forms we have considered (the typhoidand Gartner bacillus). This is the fact that it is one of thecauses of putrefaction of proteids. It forms indol, for

example, in proteid solutions ; it is found in foul water. in

oil, and it is a constant inhabitant of the intestines, beingound in that region soon after birth. After death it some-imes penetrates the different abdominal organs and is one ofhe causes of the putrefaction of the body. These charac-eristics, therefore, mark it sharply off from the twoither micro-organisms. There are, no doubt, manylifferent forms of the bacillus coli which differ inninute particulars, some, for example, forming hardlymy indol, others coagulating milk slowly. Withoutiowever, entering into the different forms which thisbacillus may assume or into a discussion as to whether theyire all modifications of the same form I. may say that thebacillus I worked with was one which was obtained from thespleen in a case of typhoid fever and which no doubt cameoriginally from the intestinal tract. The virulence of thebacillus coli as obtained from the intestines varies consider-ably and several observers have described an exaltation andan increase in the numbers of the bacillus in the courseof cases of typhoid fever. The bacillus I obtained fromthe spleen was more virulent than the stock culturesof the bacillus which I had in the laboratory andwhich were obtained from different sources. Its viru-lence was readily increased, much more readily than thatof the typhoid bacillus or of Gartner’s bacillus. 6 c.c.of a fourteen days’ old broth culture injected intothe peritoneal cavity of a large guinea-pig caused deathin under twenty-four hours ; 4 c.c. of a broth cultureof the peritoneal exudation of the first guinea-pig wasinjected into the peritoneal cavity of a second andkilled the animal in under twenty-four hours. In the

succeeding experiments 2 c.c. of the peritoneal fluid werefound to kill ; 0’5 c.c. was fatal at the fifth inoculationand at the sixth it was found that 0-25 c.c. and 0’1 c.c.killed the animals in four hours. At this stage the virulenceof the bacillus was considered sufficiently exalted for experi-ments to be performed. Throughout, the cultures obtainedfrom the peritoneal fluid were pure and always gave thetypical reactions of the bacillus coli. The results ofthe injection of the bacillus into the peritoneal cavitydiffered markedly from those obtained from the injectionof the other two bacilli, mainly in the fact that the effusionwas blood-stained, more so at first than later, when thebacillus became more virulent. The blood-staining was notdue to the exudation of corpuscles but to the colouringmatter of the blood being dissolved in the liquid, which washighly albuminous. Numerous petechise were also observed,both beneath the peritoneum and the mucous membrane ofthe ileo-csecal region and csecum. No peritonitis was observedat any time. The toxic products of the bacillus were

investigated by growing it in broth and in proteid solutions.Growth of the bacillits in broth.-An early experiment

was performed with the bacillus which was not extremelyvirulent and which was grown in broth for twenty - threedays. The broth was then filtered through a Chamberlandfilter and 8 c.c. of the sterile filtrate were injected intwo- doses on successive days into the marginal vein ofthe ear of a rabbit weighing 1090 grammes. Theinitial effect of the first injection was a fall of temperaturein about two hours which lasted over four hours, the totalfall of temperature being 5° F. The result of the secondinjection on the following day was practically nil, but on thethird day the temperature was subnormal. It then beganto rise gradually and there was some degree of fever on thefifth, sixth, and eighth days of experiment, which wasmarkedly increased on the ninth day. On the tenth daythere was a fall and a subsequent rise on the eleventh day,after which the temperature of the animal became practi-cally normal. The animal steadily lost weight during thefirst two days of experiment-100 grammes the first day and38 grammes the second day, the weight remaining stationaryon the third, fourth, and fifth days of experiment, afterwhich it slowly regained its normal. The animal exhibitedno other symptom; there was no diarrhoea and completerecovery took place. This result shows that the poisonousproducts which are formed in the broth by the bacillus colihave the initial effect of reducing the temperature of thebody as well as weight and that subsequently there is a

reactionary rise of temperature, which remains at a febrilepoint for several days.Other experiments were done with the broth cultures in

the same way as those done with the typhoid bacillus andwith Gartner’s bacillus-viz., after a period of 18 days’incubation the bacilli were killed with chloroform and the

injection of the clear broth obtained by centrifugalising and... ()

4

of the broth with the bodies of the bacilli was performed indifferent animals. In the first experiment 4 c.c. of clear brothwere injected into the marginal vein of a rabbit weighing720 grammes. In 2! hours the temperature had fallen20 and there was slight diarrhoea. In 4t hours therewas a further fall of 25° and the temperature altogetherremained subnormal for over five hours. On the next dayit had regained its normal limit, though the animal con-tinued to lose weight. In the second animal a similar doseof 4 c.c. of the broth containing the bodies of the bacilluswere injected into the marginal vein. The temperature fellabout 15° in two and a half hours, after which it began torise slightly, and during the whole of the next day thetemperature of the animal was febrile. In one and a halfhours after the injection there was slight diarrhoea and theanimal lost 70 grammes in twenty-four hours. The tempera-ture remained at about the normal level till the fifth day.On the seventh day it had fallen and there was slightdiarrhoea. On the eighth day there was a fall to 98° F. andon the ninth day when death occurred the temperature wasbelow 95°. The total loss of weight was 210 grammes.These results were compared with the results of injecting asimilar dose of the same stock of broth containing the bodiesof the bacillus but which had been subjected to heat. In thefirst of these experiments the mixture of broth and bodieswas heated at 60° C. for ten minutes. The result of the

injection was the cause in a very short time of a febrile riseof temperature which lasted the whole of the day and waspresent in the morning of the following day, the tempera-ture then gradually falling to the normal. In one and a

quarter hours after the injection there was profuse diarrhoeaand the animal lost somewhat in weight during the firsttwo days of experiment. In another experiment a similardose was given after being heated to the boiling point ofwater for ten minutes. In less than an hour after theinjection profuse diarrhoea was observed and the tempera-ture rapidly fell, the total fall being 7°F. The animal diedin six hours.

It is seen from these experiments that the clear broth-thatis, the broth not containing the bodies of the bacillus-pro-duces a fall of temperature and loss of weight with somediarrhoea and that a similar dose of the broth with the bodiesof the bacillus causes some lowering of temperature and areactionary febrile rise and that the effect of heating themixture of broth with the bodies of the bacillus is to increasethe poisonous activity of the solution, so that when themixture was heated to 100° C. it proved fatal to the animalin six hours with great lowering of temperature. The

explanation of this appears to be that the action of the heatcauses the discharge of the poison which is present in thebodies of the bacillus. 1. The type of action of the poison ofthe bacillus coli appears to be the same as that of the othertwo bacilli investigated, in some cases producing a greatfall of temperature and in other cases a rise of temperature.2. Heating the dead bodies of the bacillus suspended in thebroth culture fluid increases the toxicity of the solution, asin the case of the two other bacilli, but in the case ofthe bacillus coli it requires a temperature of the boiling pointof water to effect this, whereas with the other two micro-organisms such a temperature rather diminishes thanincreases the toxic action. Altogether, the mode of actionof the poisonous products of the bacillus coli is more

irregular than with the other two bacilli, not only as

regards its lethal but also as regards the irregular kind offever and after-fever produced.

C,’rort7z of the bacillus in proteid solutions.-As with theother bacilli, very different forms of proteid solution wereused as culture medium-liquid serum, serum diluted with saltsolution, serum diluted with broth, alkali albumin in broth,and Marmorek’s fluid. The general result obtained in theseculture media was that the bacillus coli digested the proteidspresent and produced an abundance of albumoses (a muchgreater quantity than the typhoid bacillus and Gartner’s

bacillus), so that at the end of a month’s incubation of sucha culture a brilliant biuret reaction was obtained. The

digestion of proteids by the bacillus coli is not nearly somarked as it is with the anthrax bacillus or the diphtheriabacillus. A more remarkable effect was observed in thefact that after the bacillus had been growing in a proteidsolution for some time it precipitated the proteid in theform of a clot, and this occurred more particularlyin the solutions of diluted serum and in Marmorek’s fluid.The results were compared with a control ‘flask whichwa" nut inoculated and which was kept in the lncubator

alongside the cultures. The precipitation of the proteidin the form of a gelatinous clot was not constant althoughit occurred in the majority of cases. It did not appear tobe due to any increased acidity of the culture medium pro-duced by the growth of the bacillus in it, inasmuch as nosuch increase of acidity was present, the liquid remainingalkaline. The phenomenon appears to be a special actionof the growth of the bacillus coli. In some solutionsGartner’s bacillus produced the same result. Thus fiveflasks were prepared containing about 100 c.c. of Marmorek’sfluid, which is a mixture of one part of ascitic fluid with twoparts of broth. Two flasks were inoculated with the bacilluscoli, two were inoculated with Gartner’s bacillus, and onewas kept as a control. In fourteen days a copious and softclot had formed at the bottom of all the four flasks, the liquidin the control flask remaining limpid and unclotted. Innone of the similar experiments with the typhoid bacilluswas any clot or precipitation of the proteid observed,so that this serves as another distinction between thetyphoid bacillus and the two other bacilli. The clotting ofthe proteid solution is of interest in connexion with theexperiments of Stillmark, who found that the substancecalled ricin, which is obtained from the seeds of the castor-oil plant and which -will be discussed later, also producesa clotting in proteid solutions, especially serum. Stillmarkused it as an argument in favour of the ferment nature of thetoxic body, ricin.

THE RELATIONSHIP OF SOME FORMS OFCOMBINED DEGENERATIONS OF THE

SPINAL CORD TO ONE ANOTHERAND TO ANÆMIA.

BY J. S. RISIEN RUSSELL, M.D. EDIN., F.R.C.P. LOND.,ASSISTANT PHYSICIAN TO UNIVERSITY COLLEGE HOSPITAL AND

PATHOLOGIST TO THE NATIONAL HOSPITAL FOR THEPARALYSED AND EPILEPTIC, QUEEN-SQUARE.

INTRODUCTION.

THOUGH the writings of Lichtheim 1 ten years ago firstcalled general attention to the association of symptomsindicating affection of the spinal cord with those of per-nicious ansemia, Leichtenstern in 1884 described two

cases of what was evidently the same condition, as tabes

associated with ansemia. During the decade which has elapsedsince Lichtheim’s first paper many important communica-tions on the subject have been made which have greatlyadvanced our knowledge not only of the clinical character-istics of this affection of the spinal cord but also in regardto its etiology and pathology. Among the earlier contribu-tions to the subject two emanated from the National

Hospital for the Paralysed and Epileptic, the first byDr. H. M. Bowman,s whose promising career was so earlyterminated by death, and the second by Dr. James Taylor,4whose important communication holds a leading placein the history of the disease. I feel it in a measure myduty, therefore, to record our further experience of theaffection at the hospital since the second of these papersappeared in 1895, and I am greatly indebted to Dr. CharltonBastian, Sir William Gowers, and Dr. Ferrier for allowingme to make use of the cases which form the subject of thepresent communication. Although a large number of papershave appeared on the subject of the changes found in thespinal cord in cases of pronounced ansemia, and although afew writers have suggested a possible connexion between thiscondition and that described by Gowers under the title ofataxic paraplegia, no serious attempt has been made todiscuss the relationship fully or to establish a connexionbetween the two affections, if we except an important paperby Rothmann which deals with the whole subject of the

primary combined tract affections of the spinal cord.Besides presenting other features of interest the followingcases appear to me to afford a good opportunity of discussing

1 Neurologisches Centralblatt, 1887, p. 236.2 Deutsche Medicinische Wochenschrift, 1884.

3 Brain, 1894, vol. xvii., p. 198.4 Transactions of the Royal Medical and Chirurgical Society, 1895,

vol, lxxviii., p. 151.5 Deutsche Zeitschrift für Nervenheilkunde, 1895, Band vii., p. 171.