streptococcic, or antistaphylococcic sera; in fact, all the evidence points in the opposite direction.

In treating a case of septicemia, therefore, we require a serum which is specific against that individual species and variety of pyogenic coccus which is producing the symptoms. It is manifestly absurd to inject an antistreptococcic serum into a patient suffering from a staphylococcic infection; and it is apparently equally futile to employ against one variety of streptococcus a serum prepared from a different variety of streptococcus. I must admit, however, that one or two cases of severe pyogenic infection have apparently benefited by treatment with the wrong anti-serum; and this may be explained by the fact that the injection of a serum produces, apart from any specific effect, a leucocytosis.

6

In the light of this argument, let us enquire how far antistreptococcic serum' has hitherto proved successful from a practical point of view.

Dr. Ogle analysed all the cases recorded in the Lancet and the British Medical Journal during the years 1896 to 1901 inclusive, and found 110 cases with 70 recoveries. But of these very few had had an examination of the blood made, so that the greater number are valueless. Unless a bacteriological examination has shewn organisms to be present in the blood, we are justified in saying that the case may have been one of sapræmia only, which would-or might-have recovered after efficient local treatment. How often do we see recorded "a case of puerperal septicemia successfully treated with antistreptococcic serum," and how rarely has the fundamental requirement of the case, namely a bacteriological examination of the blood, been complied with. I am far from implying that these sera are useless in cases of septic intoxication, but I am at present concerned only with septicemia.

In Dr. Ogle's series only a few had had a bacteriological examination made of the pus at the site of infection, so that in the majority of the cases it was not even ascertained that the organism causing the symptoms was a streptococcus.

In only 13 of the 110 cases was a streptococcus found in the blood. Of these, six were due to puerperal sepsis, three to middle ear ear disease, and four to other sources of infection. Ten recovered and three died, but we are not justified in assuming a general recovery rate of 77 per cent., remembering how small a proportion of unsuccessful cases ever attain the dignity of publication.

In the same paper 19 cases of malignant endocarditis are cited as having been treated with antistreptococcic serum, with six recoveries. But here again in only three of the six instances were streptococci found in the blood.

An extraordinary example of perverted reasoning is to be found in the Report of the American Gynecological Association in 1898 of their Committee appointed to investigate the value of antistreptococcic serum. In 101 cases of gynaecological sepsis, streptococci were found, either locally or in the blood, and these were treated with an antistreptococcic serum, with a mortality of 32.6 per cent. In 251 cases no bacteriological examination at all was made but the serum was employed. The death rate in this series was 15.85 per cent.

These gynæcologists solemnly concluded that the medical profession was not justified in proceeding further with the use of antistreptococcic serum. We can only hope that they did not follow out their own recommendation, and that the use of more recently produced sera, applied more scientifically, may have caused them to alter their opinion. The tendency at the present time is to produce an antiserum which, being prepared from several strains of the organism in question, shall have a better chance of effectiveness than one produced from a single strain. Such a serum is termed multivalent. Dr. Foulerton estimates that 46.3 per cent. of cases of puerperal sepsis are due to streptococci, and has produced a multivalent serum from five different strains of streptococci obtained from puerperal cases, a serum which seems to promise great results.

In a similar manner a multivalent antistaphylococcic serum can now be obtained, prepared from several strains of the staphylococcus aureus, which has been used in several cases at this hospital, but in too few instances for us to be able at present to say much about its efficacy. In a few it has apparently been employed with success, but usually empirically as the bacteriological examination has either been neglected, or has demonstrated organisms different from those against which the serum is specific.

I cannot leave the consideration of pyogenic infections without an allusion to the colon bacillus. The term includes a group of closely allied organisms, which, although normally, and possibly beneficially, present in the intestinal canal, may, under abnormal conditions, become highly pathogenic.

It has recently been demonstrated that in the great majority of cases of peritonitis of intestinal origin, it is the colon bacillus which is responsible for the fatal issue. It is also known that some infections of the urinary tract, and other parts, are due solely to this organism. At Dr. Dudgeon's instigation a well known firm have recently made a multivalent serum, prepared by rendering animals immune against seven highly pathogenic strains of colon bacillus isolated by him from cases of peritonitis, together with four other strains isolated by Dr. Dowson from other sources. This multivalent serum is both antitoxic and antibacterial. It has been used at this

hospital in several cases, in one or two with most striking results. The first time it was employed was upon a case under Dr. Turney suffering from an acute infection of the urinary tract due to a colon bacillus. This patient had for 12 days been extremely ill with high fever, rigors and other severe constitutional symptoms; there was great tenderness over the bladder, and the urine was loaded with pus from which Dr. Dudgeon isolated a colon bacillus in pure culture.

On the 13th day treatment with the antiserum was commenced; at first in doses of 10 c. c., and then in doses of 20 c. c. In 48 hours 80 c. c. had been injected, and the temperature steadily subsided, until on the 17th day of the illness it had reached the normal. At the same time the constitutional symptoms subsided, and the amount of pus in the urine diminished.

These three cases point to the conclusion that this multivalent anticoli serum may prove to be of considerable value in cases of colon bacillus infection, of which by far the most important is peritonitis of appendicular or intestinal origin.

At the present time the attitude of surgeons towards the antipyogenic sera may be described as one of mild toleration It is used as a last resort in desperate cases, and perhaps sometimes rather as a placebo than with any real conviction that it is likely to be of much service; but if a case does get well with its use, then the surgeon either places it upon a pinnacle, from which eminence it is pretty sure to be dashed by its failure in the next case; or he shrugs his shoulders and wonders whether after all the case would have recovered without it. In fact, the thing is a fetish; it is used empirically, and the result is accepted according to the degree of faith which was reposed in it.

Now this attitude is entirely wrong, and cannot be too severely criticised. These sera are produced as the outcome of delicate, accurate and laborious scientific work by the pathologist; it is for the practical therapeutist to see that his part of the work is done in an equally careful and scientific manner. The man who indiscriminately treats a disease, the exact nature of which he is ignorant, with a serum of which he is still more ignorant, is guilty of deliberately placing a drag upon the wheel of scientific progress.

I would urge that these sera have not had, and are not having, the fair trial to which they are entitled, and it is only when certain conditions shall have been fulfilled, and that in an adequate number of cases, that we shall be entitled to pass a judgment of any value upon them.

1. The cases must be carefully selected. It is obviously useless administering serum to a patient on the point of death-one whose vital powers are at so low an ebb that it cannot be absorbed and

utilised; whose moribund tissues are incapable of making any response. Dr. Dudgeon tells me that antitoxin injected into the cellular tissue of animals extremely ill from diphtheria is, after death, found to be still present at the site of inoculation.

It is equally unfair to administer it unnecessarily to a patient who would get well without. Between these two extremes lie the suitable cases; and it is unfortunate that in the nature of things their selection must to a large extent be a matter of individual opinion. Still, it is not too much to expect that, with the exercise of a moderate degree of discrimination, this source of error may be reduced to a minimum.

2. Every means of local treatment must first have been conscientiously tried-removal or disinfection of the focus of infection, or ligation of the veins in the cardiac side of the focus. It is unfair both to the patient and the treatment to inject the serum as long as a foul clot remains in the uterus, or in the lateral sinus.

3. Bacteriological examination both of the focus of infection and of the blood should be made, in order to form an accurate diagnosis, and to ascertain what organism or organisms are the cause of the disease.

4. The appropriate serum must be used. Herein lies the greatest difficulty, but it is at least possible to avoid using an antistreptotoccic serum for a staphylococcic infection, and by employing a multivalent serum, one is more likely to obtain a satisfactory result. In cases of mixed infections more than one kind of serum might advantageously be employed.

Theoretically, the best plan would be to use a serum prepared against the very organism isolated from the individual case. In most instances, however, the acuteness of the disease renders this impracticable, but there are cases of chronic septicemia in which such a proceeding would be quite possible. Some here will be able to recollect such a case-that of a man who, after a hernia operation, was, for a period extending over two or three years, a frequent visitor to the hospital for the purpose of having metastatic abscesses opened, until he eventually succumbed to lardaceous disease.

5. The dosage must be adequate.

The dosage must necessarily always be empirical, as it is impossible to estimate with accuracy the severity of the infection which is to be combated. But it is unlikely that 5 or 10 c.c. of serum, unless it be one of exceptional potency, given at intervals of many hours, will be of much value. The case which gets well with such emasculate dosage would probably recover without any serum at all. In the treatment of diphtheria some reasonable attempt is made to

suit the dosage to the severity of the case, but with the pyogenic infections that is a course which is but rarely adopted. One must of course remember that in diphtheria the antitoxin is standardised, and the dose is estimated in units, by which means the amount administered can be regulated. Such, however, is not the case with the antipyogenic sera, as ordinarily supplied, and the dose of these sera is reckoned in cubic centimetres, so that it is possible that 5 c.c. of one brand of serum may be equivalent to 10 or more c. c. of another.

Nevertheless, when we remember that on the one hand, in many cases where small quantities have been employed, a serum has proved futile; whilst, on the other hand, many successful cases have received very much larger amounts, we cannot but conclude that inadequacy of dosage is probably responsible for many of the failures.

A successful case, reported by Ballance and Low, received 263 c.c. in six days, but other recorded cases have received much larger quantities than that. Foulerton is of opinion that a fair trial has not been made unless at least 40 c.c. have been given within 12 hours. Berg advocates doses of 200 c.c. of antistreptococcic serum.

There is an impression abroad that the dosage should be regulated according to the age of the patient. In the paper by Milward, to which I have already referred, a case is alluded to where 50 c.c. of antitetanic serum were given to a child only 14 days old. This, writes the author, would be equivalent to giving to a ten stone man 583 c.c. in the same period.

There one must emphatically disagree. Providence does not adapt the dose of toxins to the weight of the patient; and it is therefore necessary for us to regulate the antidote, not by age or weight, but according to what we may judge to be the severity of the infection.

It has never been shewn that the use of reasonably large doses of antitoxin is harmful, although in excess the injection of a serum produces hæmoloysis. The rashes and joint pains which sometimes occur can be produced by the injection of fresh normal horse serum, and are both transitory and unimportant.

6. The serum must be fresh. This is a point of the greatest moment. An antibacterial serum very rapidly deteriorates, and the "immune body," present in such a serum, is useless without the

66

complement," which is a very unstable body, and only present in small amount. If the "complement" in the patient's blood is also inadequate, then the "immune body" will be unable to act. Wassermann found that by giving fresh normal ox serum, together with the specific antiserum, he could protect animals against virulent