Vedder and Clark (1912), reporting observations on polyneuritis gallinarum based on experiments lasting for several years during which time over 200 fowls were used, make the following statement with regard to the relation between the amount of polished rice eaten and the time of development of neuritis.

In April, 1917, while examining some rats (probably M. decumanus) in company with Mr L. D. Cleare, Jnr., my assistant, for external and internal parasites, the lungs of a male specimen were observed to be in a highly diseased condition.

During four months' work with anaerobes, which were derived from wounds, a number of species have been noticed which appear to be hitherto undescribed. Two of these are proteolytic organisms and are not pathogenic to guinea-pigs when administered in large doses.

Compared with 20, or even 10, years ago our present knowledge of food poisoning outbreaks is extensive and in certain directions fairly complete. In spite of this greatly extended knowledge there are some aspects in regard to which we are yet lacking in fundamental information. This is particularly the case as to the precise sources of infection. It may be accepted as a demonstrated fact that most outbreaks of food poisoning are due to infection of the food eaten with one or other member of the Gaertner group of bacilli. The present paper is only concerned with the outbreaks associated with this group of organisms. A study of the individual outbreaks usually supplies evidence which definitely incriminates a certain article of food, and for most of the recent outbreaks further evidence is forthcoming that this has been infected with one or other member of the Gaertner group of bacilli. Tracing the matter a step further back it is only in a quite small minority of outbreaks that the recorded facts show how the food has become so infected. In a proportion of cases, perhaps more than half for continental recorded outbreaks but in only a small fraction of the British outbreaks, it is true that definite evidence is forthcoming showing that the meat was derived from an animal itself suffering from general or local disease caused by Gaertner group bacilli. Even, however, for these cases our recorded knowledge ceases with this information, and we do not know how these animals became infected or whether they represent isolated cases or are part of widespread epidemics amongst the animals affected.

The examination of the spleens of 24 pigs and 10 calves and internal organs from 12 other pigs, all passed as healthy and fit for human consumption, failed to show the presence of any organisms belonging to the Gaertner group. One para-Gaertner bacillus was isolated in pure culture from the spleen of one of the calves.

The spleens of the 24 pigs and 10 calves were also examined for the presence of aerobic organisms generally. Exactly half of the pig and half of the calf spleens were sterile, the remaining 50 per cent, containing bacilli which were not completely identified but were mostly B. coli, non-lactose fermenters allied to B. coli, streptococci and staphylococci. All the organs were examined within a few hours of death. There was no correlation between interval since death and bacterial content nor any relationship to the age of the animal. The bacteria were evidently present in these organs at the time of death.

An extensive series of examinations of sera of animals passed as healthy showed that in a considerable proportion of them specific agglutinins were present against one or other member of the Gaertner group. These agglutinins were absent in all the calves, were mostly absent from sheep, but were fairly well developed in 40 to 50 per cent. of the pigs, cows and oxen examined. They were also present in a considerable proportion of the horses tested. The failure to demonstrate these agglutinins in the calf and their definite development in many of the cows and oxen sera suggest that they are not present in the new born animal but develop later in life. The available data is insufficient to enable a definite opinion to be given as to whether these positive reactions with Gaertner strains are to be ascribed to an old infection with these bacilli but they suggest this possibility.

As a practical point these results show that the demonstration of the presence of agglutinins, in moderate amount, against this group of bacilli in a suspected animal cannot be accepted in itself, and without further corroborative evidence, as proof of an existing infection with these organisms or as conclusive evidence connecting the animal with an outbreak of food poisoning. Much less so can the presence of such agglutinins be accepted as evidence warranting the condemnation of the carcase as unfit for food.

Amongst the animals examined suffering from definite disease one cow which died of septic poisoning showed the presence of a considerable amount of specific agglutinin in its serum, but the other bacteriological data was sufficient to say that the cause of death was not a Gaertner group infection.

The 48 rats examined were selected as showing no macroscopic evidence of Gaertner group infection and from none of them was a member of this group isolated, although five para-Gaertner strains were found which very closely resembled these bacteria.

In nearly every case the sera of these rats failed to show the presence of any Gaertner group agglutinins. This fact is of particular interest in view of the rather different findings which Read and I obtained with another series of rats from a different source and where the possibility of infection with Danysz virus (a Gaertner group organism) was likely. They add weight to the view that in these (mostly young) animals the presence of Gaertner group agglutinins to any considerable extent is evidence of an old infection with a member of this group, and is confirmatory evidence in favour of the agglutinin results with animals used for food being of a similar origin.

The absence of any correlation between the presence of the six para-Gaertner strains and the presence of agglutinins against true Gaertner bacilli is in favour of the view I have expressed from my earlier work, that there is no relationship between these two groups of organisms and that the para-strains cannot be considered as modified Gaertner organisms which under favourable conditions can revert to that type.

The examinations of specimens from diseased animals are too few to enable any deductions to be drawn, but I remain strongly of opinion that work along these lines is likely to throw valuable light upon the etiology of food poisoning and should be systematically undertaken by those who are in a position to obtain and examine such material.

The study of the factors influencing the heat-denaturation of serum proteins has led us to change the technique of my method (1916) for concentrating antitoxic sera. These changes involve a considerable shortening of the process and therefore a brief description of our present method of procedure may be useful to those engaged in serum concentration in other laboratories.

[Up to recent years, fish were frozen in a dry chamber; this was not found entirely satisfactory, the fish tending to become dry and lose their flavour. We were ourselves intending to experiment on freezing fish in brine prepared in different ways and made a proposal to the Fish Food and Motor Loan Committee to this effect. Our attention was then drawn to Ottesen's process which we have since tested through the kindness of Mr J. V. Pryor of Cambridge2, who has supplied us with ice and allowed us to use his cold storage chambers. The confirmatory results, which we have obtained practically, are well explained in the light of Dr Reuter's report on his careful investigations, an abstract of which follows.