1. The red blood corpuscles of different species have definite and distinct mean lytic points.

2. The apparent tonicity of sheep serum has no relation to the mean lytic point of the red corpuscles of that individual.

3. The apparent tonicity of sheep serum is not due to its proteins but to its lipoids.

4. The residue of sheep serum, after removal of proteins and lipoids, is sufficiently alkaline to cause lysis.

(1) If a series of dilutions of an antiserum are prepared, a dilution can be selected which when mixed with an appropriate quantity of the homologous antigen forms no precipitate but nevertheless binds complement. If complement is added to such a mixture of antigen and antiserum a precipitate appears after an interval of six to twenty-four hours.

(2) By using an appropriate mixture of antigen and antiserum the amount of the precipitate can be increased, within certain limits, by increasing the amount of guinea-pig serum (complement) present in the mixture.

(3) Similar results can be obtained if an euglobulin solution prepared from guinea-pig serum is substituted for the normal guinea-pig serum.

(4) A suitable mixture of antigen and antiserum precipitates the euglobulin of guinea-pig serum in a manner which may be compared with the precipitation of euglobulin by carbon dioxide.

(5) By keeping a, precipitating mixture of antigen, antiserum and complement at a temperature of 0° C. it is possible to demonstrate that the resulting precipitate contains the mid-piece fraction of the complement in an active state.

The mid-piece fraction is not used up when it is bound by a mixture of serum with its homologous antiserum.

(6) The precipitate which results from the interaction of antigen and antiserum fixes both fractions of the complement. The mid-piece is fixed much more readily than the end-piece but it is possible to demonstrate a stage when a considerable quantity of end-piece has been bound while a small quantity of mid-piece still remains free.

(7) The results obtained in these experiments show that the fixation of the fractions of the complement by a mixture of antigen and antiserum is essentially similar to the fixation produced by suspensions of barium sulphate and similar complement fixing substances. These experiments confirm the recent work of Gengou on this subject.

(8) The particles of a precipitate probably adsorb the euglobulin of the guinea-pig serum and this adsorption of euglobulin is an essential part of the mechanism of complement fixation.

1. Over 450 naturally infected or wild flies (Musca domestica) were caught in Liverpool during September and the first part of October 1911 from different parts of the city. The number and kinds of bacteria carried and contained by them have been investigated.

2. The number of bacteria coming from house-flies whilst struggling in liquid may be very large, varying from 2000, the lowest figure in 5 minutes, to 350,000, the highest figure in 30 minutes. This number may be taken as a measure of their capacity to pollute liquid with their vomit or excrement, or by wallowing in it. The number of bacteria carried inside the fly is very much greater.

3. Flies caught either in insanitary or congested areas of the city carry and contain far more bacteria than those from the more sanitary, less congested or suburban areas. The number of aerobic bacteria from the former varied from 800,000 to 500,000,000 per fly, and from the latter from 21,000 to 100,000.

4. The number of intestinal bacteria as indicated by glucose bile salt fermenters is also greater in the insanitary or congested areas, the numbers varying from 10,000 to 333,000,000, than in the more sanitary areas where they carried from 100 to 10,000.

5. Pathogenic bacteria and those allied to the food poisoning group were only obtained from the congested or moderately congested areas and never from the suburban areas.

6. We have examined the morphological characters and cultural reactions of 123 strains. Among those identified were two Streptococci, and several Staphylococci and Sarcinae. 106 were small Gram negative non-spore bearing bacilli, and have been grouped as follows:

Chromogenic group. Two strains of B. pyocyaneus were isolated from a knacker's yard; for the first time, we believe, from wild flies.

Colon group. 41 colonies of this group were picked off haphazard and classified according to McConkey as follows:

B. acidi lactici type 19·5%

B. coli communis type 12·2 %

B. neapolitanus type 19·5 %

B. lactis aerogenes type 46·4 %

Salmonella group. One bacillus gave identical reactions to B. enteritidis of Gaertner except that serological tests were negative.

Morgan's infantile diarrhoea group. One identical to Morgan's No. 1, and many others closely resembling it and Morgan's Nos. 2 and 3 were obtained.

Others fall into Proteolytic, Acid lactose-sucrose (saccharose), and Miscellaneous groups.

7. Flies caught in milk shops apparently carry and contain more bacteria than those from other shops with exposed food in a similar neighbourhood. The reason of this is probably because milk when accessible, especially in the summer months, is suitable culture medium for bacteria, and the flies first inoculate the milk and later reinoculate themselves, and then more of the milk, so establishing a vicious circle.

8. On one occasion we compared the number of bacteria carried by house-flies caught in an eating house opposite the slaughter houses with the number carried by blue-bottles; the latter, as might be expected, was far larger.

9. In cities where food is plentiful flies rarely migrate from the localities in which they are bred, and consequently the number of bacteria they carry depends upon the general standard of cleanliness in that locality. This is well indicated by the fact that flies caught in a street of modern fairly high class workmens' dwellings forming a sanitary oasis (Hornby Street) in the midst of a slum area, carried far less bacteria than those caught in the adjacent neighbourhood.

10. It is clear that flies from the suburbs where infantile diarrhoea is rare carry far less bacteria than those in the city where it is common. It was, nevertheless, impossible in the time at our disposal to correlate exactly the number or varieties of bacteria carried by flies in the city with the number of cases and deaths from infantile diarrhoea in individual streets.

11. As the amount of dirt carried by flies in any particular locality, measured in terms of bacteria, bears a definite relation to the habits of the people and the state of the streets, it demonstrates the necessity of efficient municipal and domestic cleanliness, if the food of the inhabitants is to escape pollution, not only with harmless but also with occasional pathogenic bacteria.

The two processes involved in the phagocytosis of bacilli (B. typhosus) by polymorphonuclear leucocytes in the presence of a sensitising fluid, viz. (1) the sensitisation of the micro-organism, and (2) the phagocytosis of the sensitised micro-organism, have been separately investigated. A single guinea-pig end-piece acted as the sensitising fluid throughout all the experiments.

The removal of opsonin by bacillary emulsions was found to run the course of an adsorption process as evidenced by the numerical relations found to subsist between the bound and free opsonin in equilibrium.

Also the actual phagocytosis of sensitised or partially sensitised micro-organisms suspended in a neutral fluid (physiological saline) was found, on analysis by similar numerical methods, to proceed as an adsorption process.

It appears to be widely believed that the fever of infective disease is due to absorption of phyrogenetic substances liberated from dead microorganisms by a combuined process of lysis and extraction in the body of the infected subject. The diffusion of products of the action of bacterial autolytic enzymes on microorganisms after their destruction by the defensive agencies of the body is a variant on the same theme. Closely allied are the theories of protein fever, and of anaphylactic fever, in so far as they relate to the supposed possession by microorganisms, or their split products, of pyrogenetic properties. The absorption of alien protein, or of derivatives of alien protoplasm, from organisms that are dead, is, in short, the essential feature of this conception of fever production.