disease-producing germs, human beings are more dangerous than cows, and a medical supervision of the employees of the milk plant is desirable. This is particularly important in the case of milk pasteurized in bulk, since this process gives no protection from the later contamination, which is a more or less remote possibility during the cooling and bottling processes.

The milk as it comes from the producer usually is and always should be clean. The problem of the distributor is to preserve this cleanliness.

The keeping quality of milk is more largely within the control of the distributor than is usually supposed. He is frequently responsible for the washing and steaming of the milk cans. Where this steaming is done in a perfunctory manner, particularly where tightfitting covers are applied to wet cans in warm weather, these cans become one of the most important factors in reducing the keeping quality of the milk. Where a proper washing of the cans is followed by a thoro steaming and the cans are carefully dried before being covered, they will have little objectionable effect upon the milk. The large germ content added to milk by utensils within the distributor's plant is frequently an important factor in impairing its keeping quality. The milk coolers and the bottling machines require special watching in this connection, not only because they frequently add large numbers of germs, but especially because they add them after the milk has been pasteurized.

The intelligent application of steam to all of the utensils should be a routine procedure, and the flushing out of all utensils with scalding water shortly before using them is a valuable additional precaution.

RESPONSIBILITY OF THE CONSUMER.

The food value of the milk furnished the consumer will depend primarily upon what the consumer desires and is willing to pay for. A considerable proportion of the consumers are desirous of obtaining a milk carrying 4 per ct. of fat or more, and where the milk has been sold regardless of food value they have striven to find the richest milk available at a given price. If each bottle of milk carried a statement of its fat content, the responsibility would then be upon the consumer to recognize and pay for increased food value.

The consumer thru his agents, the health officials, must determine how the healthfulness of the food supply shall be safeguarded. The dangers which naturally surround milk production and handling are such that if the milk supply is to be safe it must be protected either by a medical supervision of the health of the cows and the men or by proper pasteurization, or better, by a combination of both means of protection.

The cleanliness of milk, as it is now delivered to the consumer is in general very satisfactory, but continued emphasis is needed to insure that it shall be constantly maintained at a high level.

The keeping quality of milk is constantly receiving the attention of the consumer, since there is no other shortcoming of milk which is more quickly impressed upon him. The delivery to the consumer of old, stale milk, on the verge of souring, is quite as much a fraud as the delivery of milk deficient in food value, healthfulness, or cleanliThe consumer is constantly insisting on improvement in keeping quality, and his desires will be met as rapidly as the producer and the distributor find economical means of insuring this improvement.

ness.

Since the keeping quality of milk after delivery is dependent primarily upon the temperature at which the milk is held, the responsibility rests upon the consumer to hold the milk at a reasonably low temperature after it is delivered to him. Too frequently little regard is given to this matter by the consumer, and much of the criticism directed against the keeping quality of milk is accordingly unjust.

PREFACE.

The present bulletin will appeal to the average reader as a technical discussion of analytical methods of no immediate interest to him. For that reason this preface is written.

During the past twenty to twenty-five years much has been written in a popular way about the number of bacteria in milk until the general public has come to have the feeling that the average glass of milk is to be looked upon with suspicion. Because of the difficulties involved in making a simple explanation of the methods used in making counts of objects as tiny and as numerous as bacteria, it has become a general custom among dairy bacteriologists to speak of the counts which have been made as counts of the number of bacteria present in milk. At the same time it has been recognized by these men that counts made by the methods in use probably represented counts of "groups" of bacteria rather than counts of single bacteria. Occasionally someone has raised his voice in protest against the general custom of referring to the counts as counts of bacteria; but the protest has been just as frequently ignored.

The development of the new microscopic technique described in a bulletin issued by this Station in 1914 has now made it possible to make reasonably accurate counts of the individual bacteria in milk. The result has been to show that the number of bacteria in milk is much greater than the counts commonly given would indicate. At the same time it has been definitely shown that the chief reason why counts made by the generally used plate method are too small is because they are really counts of masses of bacteria rather than counts of individual bacteria.

It will be a surprise to those who have not realized the probability that this is the case to learn that milk commonly stated to contain 10,000 bacteria per c. c. rarely has less than 15,000 bacteria and may have 250,000 or even more per c. c. Similar conditions' hold in regard to other figures that are commonly given.

The confusion which is sure to arise because of the unfortunate situation caused by the previous inaccurate statements can now only be met by care on the part of each person to avoid such inaccuracies in the future.

For the sake of those who may be inclined to look with still greater suspicion on milk when they learn that it contains many

more bacteria than they had supposed, it should be stated that recent work in agricultural bacteriology has emphasized the fact that harmful bacteria are the exception rather than the rule. Both sour milk drinks and cheese contain many more bacteria than are present in milk and yet they are known to be healthful foods. The human intestine is known to harbor countless myriads of these tiny organisms in all cases; and yet their presence in the body causes no injury. In fact, such experimental evidence as is available indicates that their presence is an actual benefit. It is only when disease-causing organisms are present that troubles follow.

For the sake of those who, because of selfish or other reasons, would like to conclude that bacteria counts are valueless because they are subject to gross errors, a simple comparison will suffice. In looking at a distant hillside it is not possible to count the trees thereon. Nevertheless it is ordinarily quite possible to form an accurate judgment as to the area covered by woods, by wooded pasture land, and by meadows. Such a judgment might have just as great value for the purpose in hand as would one based upon an accurate count of the trees per acre of land. Bacteria counts of market milk and cream have their value regardless of the difficulties which prevent them from being absolutely accurate. The studies which have been made at this Station and elsewhere upon the sources of bacteria in milk have made it increasingly plain that the correlation between the number of bacteria in milk, and the presence or absence of disease-causing organisms, or of the presence or absence of dirt, dust and filth, is so slight that bacteria. counts have little significance with regard to either of these conditions. There remains however a close correlation between the number of bacteria and the state of freshness of the milk or its keeping quality. Milk containing few bacteria ordinarily has a much higher keeping quality than milk containing many bacteria, and the housewife who supposes that she is buying fresh milk which will not sour within a reasonable time should be protected by an efficient inspection service against the milkman who sells her a quart of milk which is nearly sour.

The universal need of inspection in order to control the quality of milk makes studies concerned with improvements or simplifications of the methods of counting bacteria in milk of general interest.

ROBERT S. BREED.

THE NUMBER OF BACTERIA IN MILK.*

JAMES D. BREW AND W. D. DOTTERRER.

SUMMARY.

1. Comparative counts of the bacteria in 643 samples of unpasteurized market milk were made by the direct microscopic method and the plate method. The counts under the microscope were made in two ways, namely, counts of individual bacteria and counts of groups " of bacteria.

66

2. These counts have been arranged in three classes: First, 175 samples (27 per ct.) in which the plate counts were higher than the individual counts. Second, 123 samples (19 per ct.) in which the plate counts were less than the group" count. Third, 345 samples (54 per ct.) in which the plate counts were greater than the group" counts and less than the individual counts. The last class of counts are of the type which would be expected in case there were no gross errors in any of the counts.

66

3. Nearly all of the first class occurred among milk samples containing few bacteria. It seems probable that, because of the small quantity of milk examined under the microscope, bacteria were overlooked. Some of these discrepancies seem to have been caused by the presence of colonies on the plates which developed from unrecognized contaminations.

4. The low plate counts of the second class are due, either to the presence of living bacteria which do not grow on agar, or to the presence of dead bacteria in a stainable condition. Apparently the presence of the former is a more important source of error than the presence of the latter.

5. Generally speaking, the "groups" of bacteria appear to be of small average size in milks containing very few bacteria and to increase in average size as the number of bacteria increase, reaching a maximum average size of 12 to 14 individuals in milk containing less than 1,000,000 bacteria per c. c. As the number of bacteria becomes greater than this the average size of the groups diminishes. Apparently this decrease in the average size of the groups" in * Reprint of Bulletin No. 439, November, 1917.