individual schedules to all physicians, the results of which were employed to supplement the enumerators' returns. Even so, however, the figures were far below the truth, although it is probable that the results were better than those obtained at later censuses. The registration returns were, of course, far too restricted in the area and character of population included to be considered at all representative of the general mortality of the country.

GROWTH OF REGISTRATION AREA FOR DEATHS: 1880 to 1914.

1914

1913

1912

1911

1910

1909

1908

1907

1906

1905

1904

PER CENT

10

20

30

40

60

60

70

The Eleventh Census (1890) indicated a considerable growth of registration territory. The states of Connecticut, Delaware, New Hampshire, New York, Rhode Island, and Vermont, together with many cities in nonregistration states, were added, raising the percentage of total population included to 31.4. The New England group was now complete with the exception of Maine, 86.4 per cent of the total population being included. The Middle Atlantic division came next, New York and New Jersey but not Pennsylvania, with a per- 1900 centage of 71.5. No other geographic division had as much as one-fourth of its population included except the Pacific (25.6 per cent). The total number of deaths obtained from registration sources was 386,212, or 19.6 per 1,000 population, while the enumerators' returns furnished 455,207, or 10.6 per 1,000 of population represented. The latter figure would seem to indicate that the enumeration of deaths in nonregistration territory was considerably more defective than at the preceding census, although this may be due in part to the transfer of registration states and cities from the nonregistration to the registration area, assuming that enumerators may to some extent have supplemented their returns from registration records available. It is, of course, possible that the year 1890

may actually have been a year of lower mortality than 1880, but the known imperfections of the returns under the methods then available cast doubt upon all comparisons.

For the Twelfth Census (1900), the last for which mortality statistics were collected from both registration and nonregistration sources, the most notable feature is the beginning of effective registration in the West. Michigan was added for the census year ending May 31, 1900, and Maine completed the roll of the New England states. The aggregate population embraced for the census year was 28,807,269, or 37.9 per cent of the total. One state (Delaware) included in the previous census was omitted, it being found that the completeness of registration was below the standard.

Beginning with 1900, the reports of the permanent Census Bureau, which was constituted by act of Congress approved March 6, 1902, have related to calendar years. For the year 1900, as shown in the cartogram, Indiana as well as Michigan is included in the Middle West the first outposts of the rapid extension of better registration of vital statistics which followed in the succeeding decades. Heretofore there has been no central office of the Federal Government constantly engaged in the work. The organization of the Census Office was entirely disrupted in the interdecennial intervals, a solitary census clerk being left in charge of the records in the Department of the Interior. When the work of each decennial enumeration was undertaken, the pressing requirements attending the sudden organization of a vast force precluded the effort to promote legislation and to improve methods. Such work was begun at once, however, by Mr. William A. King, chief statistician for vital statistics, upon the passage of the permanent census law, and the result is seen in the large additions to the registration area which have been made since 'that time.

For the census year 1900 there were returned 512,669 deaths from registration records, nearly as many as those received from nonregistration sources (526,425). The death rates were 17.8 and 11.2 per 1,000, respectively; the latter is, of course, far too low, although it is probable that the enumeration was a little better than in 1890. The returns for the calendar year 1900, which were exclusively from registration records and in part overlapping with those of the census year, numbered 539,939, thus exceeding the maximum ever obtained by enumeration. No enumeration of deaths has been conducted since 1900 and, I trust, the method has forever gone into disuse except perhaps as a check upon the completeness of registration.

The Thirteenth Census (1910) omitted the mortality schedules from the blanks sent to enumerators and relied entirely upon the annual reports based upon transcripts of deaths from the registration area. As shown by the cartogram for the calendar year 1910,

this area has widely expanded. Pennsylvania (added for 1906) and Ohio (1909) now linked the registration states of the East with those of the Middle West; Wisconsin (1908) and Minnesota (1910) extended it to the great Northwest beyond the Great Lakes; Colorado (1906), Montana (1910), and Utah (1910) contributed data from the Rocky Mountain region; and California (1906) and Washington (1908) covered the long line of the Pacific coast with the exception of Oregon, which state has, during the present year (1915) adopted the Model Law. An important addition also was Maryland (1906), the first state with a considerable proportion of colored population to be included. The total population of the registration area for 1910 was 53,843,896, or nearly three-fifths (58.3 per cent) of the entire population of the country. The total number of deaths returned for the year was 805,412, or 15 per 1,000. The registration area, of which the population in 1908 comprised for the first time more than half (52.5 per cent) of the total inhabitants of the country, is becoming more and more worthy of acceptance as expressing the general condition of the Nation through its vital statistics.

Progress since the last census has continued, as shown by the cartogram (p. 11) giving the states included for 1915. Actual accomplishment has begun in the South, that very important section of the country heretofore unrepresented save by municipal statistics. Kentucky and Missouri (both added for 1911), Virginia (1913), and Kansas (1914) make good a line of state registration which extends, with the exception of the slight break in Nevada, entirely across the continent from the Virginia capes to the Golden Gate. North Carolina has also been represented in the registration area since 1910 by returns derived from municipalities over 1,000 population under a state law, which has since been extended. (1916) to include the entire state.

The condition of the registration area for deaths, encouraging as it is in comparison with former years, is still far from satisfactory. Almost the entire South, with the exception of Kentucky and Virginia, is still unrepresented by complete state returns. The quality of the returns, moreover, from certain of the states recently admitted to the registration area is still open to some question. Phenomenally low death rates suggest that the full standard of efficiency may not have been reached or perhaps maintained. Such as it is, however, it stands for a vast deal of enthusiastic effort. Many pamphlets calling attention to the importance of vital statistics and to the necessary methods for obtaining satisfactory registration have been prepared and distributed by the Bureau of the Census and by the American Medical Association, which organization has taken an especially active part in the revision and recommendation of the "Model Law," a draft of legislation which has been largely instrumental in extending the area of satisfactory registration.

In the cartograms on the opposite page, which should be examined in connection with that showing the actual registration states in 1915, those states are indicated which appear to have adopted adequate laws; that is to say, laws which, if reasonably provided with funds for execution and with competent administrators, should speedily accomplish excellent results and bring the states into the registration area. The Model Law has recently been adopted, either in full or as rules and regulations under a briefer general law, in almost the entire southeastern portion of the country, excepting only Alabama. The last two states in the country which, up to 1914, possessed no vitalstatistics laws for the state as a whole were Georgia and South Carolina. Both adopted laws last year (1914). The South Carolina law was not signed by the governor until September 1, 1914, and went into effect January 1, 1915. It has already given most encouraging results. The Georgia law unfortunately carried no appropriation, so that this deficiency will have to be supplied before it goes into actual effect. Florida adopted the complete Model Law this year (1915) and is putting it into execution. The only general section of the country now remaining without adequate legislation is the Southwest, consisting of the four states of Arizona, New Mexico, Oklahoma, and Texas. All these states have laws involving the attempt to collect vital statistics through county officials, a method which has failed to give satisfactory results.

The condition of the country with respect to the enactment of satisfactory birth registration laws is apparently about the same as that for deaths. The actual condition of birth registration is probably, however, considerably behind that of death registration, depending upon the great difficulty which is found to exist in actual practice in securing complete returns of births even under good laws in our oldest registration states.

THE REGISTRATION AREA FOR BIRTHS.

The "registration area for births" is as yet nonexistent. We have seen in the preceding section and in the cartograms showing the existence of satisfactory laws for the registration of deaths and births that a large number of the states have supposedly adequate laws for birth registration. These laws, however, are seldom, if ever, thoroughly enforced. Indeed, we may employ even more definite language and say that a birth registration law has never been thoroughly enforced on a state-wide basis in the history of the United States-with perhaps the single exception of Pennsylvania, in which state hundreds of actual prosecutions have been directed by the state authorities with good results. In practically all other states the enforcement of birth registration laws, beyond the point of gentle remonstrance, is the merest farce. No physician with present or

30026°-16-2

1 See note on p. 7.