It is at once evident that there is a general resemblance between blood serum and milk so far as this phenomenon is concerned. It is also plain that blood has a much quicker and stronger action than milk.

The results of the bacterial counts upon agar plates were confirmed by microscopical examination of stained smear preparations. At first the organisms were well distributed throughout the serum, whether heated or unheated. There were no clumps of over six or eight individuals.

At the end of six hours no organisms could be found under the microscope in preparations made from the unheated serum planted with B. lactis aerogenes. Only occasionally could the typhoid bacillus be discovered in the corresponding serum at the end of six hours. This agrees with the number of colonies upon the agar plates.

The heated serums gave quite a different picture under the microscope. Many organisms were found, lying singly, in small and long chains, and in dense clusters. This corresponded to the innumerable growth upon the agar plates.

RELATION TO PHAGOCYTOSIS.

Milk contains many polymorphonuclear leucocytes and it therefore seems reasonable to assume that active phagocytosis takes place in the fresh raw product. A priori it seemed to us that this might account for the germicidal property of milk. This assumption was apparently confirmed when we found that stained smear preparations showed but few if any bacteria in the cells in the fresh milk just after inoculation with bacterial cultures, while similar preparations made from the same milk eight hours later, kept at 37° C., showed numerous bacteria in some of the cells.

The following experiments, however, demonstrate that the germicidal power of milk is independent of its cellular contents. The

24907--Bull. 41-08-30

leucocyte-free milk is quite as active as the leucocyte-rich sediment obtained by centrifugation.

TABLE NO. 15.--Milk from a healthy cow (No. 2).

[Was centrifuged for 20 minutes at 1,500 revolutions per minute. Part of the supernatant fluid was passed through a Berkefeld filter and a clear bluish serum obtained. Five sets of tubes were inoculated with twenty-four hour agar cultures, (1) the filtered clear serum, (2) the supernatant fluid free from leucocytes, (3) the sediment rich in leucocytes, (4) the original whole milk, and (5) sterilized milk. The inoculations were made 3 hours after milking.]

This series also shows the effect of shaking the milk in breaking up bacterial clusters.

This milk was obtained from a healthy cow (No. 2) in a cleanly manner, but without special precautions. The milk was brought to the laboratory and inoculated with the cultures two hours after milking. The times given, however, are from the time of inoculation. When obtained, the whole milk contained 1,200 bacteria per

cubic centimeter. The sediment, rich in leucocytes, contained 630 bacteria per cubic centimeter, and the milk serum 490.

The milk was revolved in a centrifuge at 1,500 revolutions for twenty minutes; the sediment pipetted off was rich in leucocytes. The supernatant serum was drawn off and filtered three times through asbestos and was therefore free of leucocytes. The milk was inoculated with a pure culture of B. lactis aerogenes and B. typhosus. Four test tubes were prepared with each culture, (1) whole milk, (2) the sediment rich in leucocytes, (3) the milk serum containing no leucocytes, and (4) sterilized whole milk used as a control. The table gives the number of bacteria per loop on agar, grown at 37° C. and counted after twenty-four hours.

These tables eliminate the leucocytes and phagocytosis as a factor in the problem we are studying.

IS THE GERMICIDAL ACTION SPECIFIC ?

Almost all those who have studied this part of the problem conclude that the germicidal action of milk is specific. The following experiment confirms these observations. Here we have the same milk showing an active power of restraining the growth of typhoid and staphylococcus pyogenes aureus, but not paratyphoid A or B. Reference to our work upon the influence of heat (vide infra) upon this property of milk also indicates its specific nature.

The table again demonstrates that agglutination plays at least some part in the apparent decrease.

TABLE NO. 16.-Milk from healthy cow (No. 2).

[Inoculated 14 hours after milking. Controls, same milk heated to boiling. Organisms from 24-hour agar culture, using suspension in water of condensation in test tube.]

THE EFFECT OF DILUTION.

Whole raw milk from a healthy cow (No. 2) was diluted in the proportion of 1 volume of milk to 3 volumes of distilled water. Inoculations were made two hours after milking. Twenty-four-hour agar cultures used.

TABLE NO. 18.- Milk from a healthy cow (No. 2), inoculated 14 hours after milking. [Controls, same milk heated to 100° C. in Arnold sterilizer for 10 minutes. Organisms from 24-hour agar cultures, suspended in normal salt solution.]

B. lactis aerogenes in raw milk diluted with same milk boiled (raw 3 parts, boiled 1 part).......

B. lactis aerogenes in same milk, undiluted

B. lactis aerogenes (control, i. e., boiled milk.)...

B. typhosus in raw milk diluted with boiled milk as above (3+1).

Staphylococcus pyogenes aureus in raw milk diluted with boiled milk as

Dilution has an appreciable effect. For instance, in Table No. 17 there was a reduction of 96.6 per cent in the number of colonies of typhoid in whole milk, but only 73 per cent in the diluted milk. A similar effect is shown in relation to B. lactis aerogenes.

These experiments were originally designed to demonstrate the presence of a germicidal substance in solution, but the results again show that the action is in part due to agglutination. As far as the restraining action is concerned, the results resemble a feeble antiseptic rather than a germicide.

THE EFFECT OF HEATING AND FREEZING.

The effect of heat varies with the micro-organism. The power to restrain B. lactis aerogenes is weakened and destroyed before that for typhoid. Thus, 55° C. for thirty minutes or 60° C. for twenty minutes markedly alters or destroys the power of milk to restrain the B. lactis aerogenes, while this degree of heat shows little influence as far as B. typhosus is concerned.

Freezing milk for ten minutes apparently does not influence its germicidal power. Freezing the milk for forty-eight hours before inoculating it has no influence upon its restraining action as far as the typhoid organism is concerned, but destroys this power for B. lactis aerogenes.

In the freezing experiments a small quantity of the milk was frozen solid by a mixture of salt and ice and kept in this condition for the times stated. It was then cautiously thawed and inoculated with the micro-organisms to be tested.

The reason for selecting the temperatures used in Table No. 20, viz, 55° C. and 65° C., was to differentiate the true bactericidal from the agglutinating substances, based upon similar work upon blood serum. TABLE NO. 19.-- Milk from a healthy cow (No. 2) inoculated 2 hours after milking.