Only in camps Pike and Lewis, both National Army camps, did scarlet fever reach sufficient proportions to be regarded as an epidemic. Camp Pike leads all other camps strikingly. There were 111 admissions in November and 164 in December. The spread of the disease in this instance was in part due to its mild type, which rendered its diagnosis difficult, and in part due to inefficient quarantine measures. The men were quartered in squad rooms containing 120 men, which made it difficult to segregate the infected and exposed individuals. A small epidemic occurred also in Camp Lewis. Eleven admissions were recorded in November and 80 in December. but no deaths occurred. The spread of the disease in this camp was in part due to cross infection in the hospitals.

The distribution of scarlet fever by State of nativity has no obvious significance. However, the camps having the lowest admission rates quarter men coming from the more densely populated States. But a number of the more sparsely populated Southern States were among those having very low admission rates.

Scarlet fever as a primary cause was one of the lesser causes of deaths in 1917. As in the case of many other diseases, however, it served in many cases to lower the resistance of the patient to other and complicating disease, the various complications after proving fatal.

Of the 48 deaths attributable to scarlet fever in the United States 34 were complicated by other diseases as follows:

TABLE NO. 115.-Complications with fatal cases of scarlet fever.

Here again broncho and lobar pneumonia are the most important complicating diseases.

· VI. TUBERCULOSIS.

Though tuberculosis heads the list of causes of death by disease in the civilian population, in an Army that has been recently examined for it and from which it has been largely eliminated, naturally it will not be a disease of major importance. Yet, despite all precautions, there were in the United States troops during 1917, 7,274 admissions for tuberculosis of the lungs, giving a mean annual rate among whites of 12.52 and among colored of 33.92 (Table 116). Here, as elsewhere, the morbidity rate for tuberculosis is much higher among colored than white troops. The admission rate for white troops in Europe was relatively low in 1917, namely, 3.54. For some reason that is not quite clear the death rate of the National Guard was 50 per cent greater than that of the National Army; but this may be merely a statistical result, owing to the different ways in which the strength of the National Guard and the National Army respectively are distributed through the months.

TABLE NO. 116.-Tuberculosis of lungs, miliary, acute, and others, United States and Europe.

Tuberculosis constituted a cause of 4.1 per cent of all deaths from disease in the Army enlisted men in 1917. For the corresponding ages of life in the registration area of the United States tuberculosis constituted 34.7 per cent of all deaths. The contrast between 4.1 and 34.7 per cent indicates in a rough way the efficiency of the selective physical examination.

The number of admissions varied a good deal in soldiers born in the various States (Table 117). This is of course partly dependent upon the number of troops derived from the different States and partly upon the conditions (of the constitution of the population, or other) in each State. Assuming that the number of natives of each State is roughly proportioned to the male population, 20 to 29 years, census of 1890, the ratio of admissions to such male population has been calculated and the States arranged in order thereof (Table 117). There are a good many objections to this procedure (such as, that the quota going into the Army from the various States was not strictly proportioned to the population, etc.), but it constitutes a rough attempt to get rid of the varied strength of the Army from different States of birth.

resistance to tuberculosis of the population of States is an attribute that persists for at least a generation.

The admission rate from tuberculosis varied greatly by camps in proportion to their mean strength. Thus Camp Kearney, Cal., led in this respect (Table 49). This is readily explained in view of the susceptibility of the men who are native to the States from which Camp Kearney draws, including New Mexico, Arizona, and California. These are States that stand first, second, and fourth, respectively, in mortality from tuberculosis in civil life. They are States which have long been resorted to by victims of chronic pulmonary tuberculosis. Their descendants probably carry a tubercular diathesis. Camp MacArthur, Tex., came second. This drew from Wisconsin and Michigan, States which have not an exceptional amount of tuberculosis. Among camps with a low rate of admission from tuberculosis were Upton, 11.15; Wadsworth, 7.18; and Dix, 8.82; which drew from prevailingly urban populations.

TABLE NO. 117.-Admissions for tuberculosis by State of nativity with ratio to civil population 20 to 29 years (1890 census).

TABLE NO. 118.-Admissions from tuberculosis, by camps, with ratios per 1,000 mean strength, United States enlisted men, for September, October, November, and December,

1917.

TABLE NO. 119.-Admissions from tuberculosis by State of occurrence (excluding camps), with ratios per 1,000 mean strength, United States.enlisted men, 1917.