A canvass by the ordinary enumerators of population alone will not succeed in ascertaining the numbers of the several classes, and will totally fail of obtaining those facts relating to their condition which are essential to anything like a just view of the subject. At none of the three censuses taken under the act of 1850 have the numbers of a single one of these classes been accurately determined. In respect to some, not even an approximation was afforded. It has been exceedingly difficult for the most highlytrained specialists to draw any valuable deductions from the partial and fragmentary data obtained, while the legislator and administrator and the public generally were likely to be misled, rather than instructed, by the figures contained in the census tables devoted to these classes of the population. In this view, both of the importance of the subject involved and of the insufficiency of the agencies hitherto employed, advantage has been taken of the provisions of the act of 1879 to initiate a systematic investigation, under expert direction, of the whole field of the defective, dependent, and delinquent classes. Mr. Fred. H. Wines, for ten years the secretary of the Illinois Board of Commissioners of Public Charities, has been appointed the special agent of the Census Office, and has undertaken an inquiry which for breadth of plan and fullness of detail leaves nothing to be desired. If a moderate success be realized, of which I feel confident, the statistics will be far in advance of anything secured at any preceding census. If this scheme can be carried out to a complete result, of which I have hopes, the information obtained will have a value which it would be difficult to express. V. THE SOCIAL STATISTICS OF CITIES. In 1790 one-thirtieth of the population of the United States lived in cities of 8,000 inhabitants and over; in 1800, one twenty-fifth; in 1810 and also in 1820, one-twentieth; in 1830, one-sixteenth; in 1840, one-twelfth; in 1850, one-eighth; in 1860, one-sixth; and in 1870, a little over one-fifth. At the last date the inhabitants of cities numbered in all 8,071,875. It is probable that not only the absolute number but the proportion of the total population resident in cities will be found in 1880 to have still further increased. It will not be surprising if 12,000,000 of persons, constituting a full quarter of the population, are found living in cities of 8,000 inhabitants and over. The fact that such vast numbers are brought within limited areas not only offers an opportunity for pursuing statistical inquiries which it would be very difficult if not impossible to extend over the whole country, but it also creates a legitimate demand for additional information respecting such communities, inasmuch as they are, by the nature of city life, made subject to vital conditions widely different from those of the population generally. Moreover, the very existence of a city indicates the presence of manufacturing and commercial enterprises, which especially require careful and technical treatment in a census. For all these reasons it has been deemed best to constitute a department of the census which should be charged with collecting and co-ordinating the social statistics of cities, including all subjects proper to the inquiry. The appointment of special agent in this department has been accepted by Colonel George E. Waring, jr., of Rhode Island, and a large amount of material of a wide range has already been collected and is in process of reduction. VI.-STATISTICS OF SPECIAL BRANCHES OF MANUFACTURE. The manufacturing statistics of the census have not been subject to such overwhelming condemnation as was visited upon the statistics of mining and the fisheries, but they have never been above severe criticism on account of the inadequacy and often in a high degree the inaccuracy of the returns. Of the manufacturing establishments returned at the census of 1870, the state erroneous that correspondence was required before they could be taken up for tabulation. The labor and expense of such a service were enormous, and yet there remained at the last the reasonable suspicion that errors, not large enough to be detected with certainty in the individual, might have reached a height in the aggregate to cause a serious departure from the facts of the case. In the inquiry, what branches of manufacture should be taken up at the Tenth Census for special investigation, the question of cost has necessarily been the determining consideration. There is no branch of manufacture the statistics of which would not be improved by a special canvass, but there is a wide difference between the various branches as to the degree of improvement which might thus be effected, and also as to the expense of such a service. The following is the list of special agents appointed for this purpose: George William Bond, esq., of Boston: The Manufactures of Wool. Professor J. S. Newberry, of New York: The Building Stones of the United States, and the Quarrying Industry. James M. Swank, esq., of Philadelphia: The Manufactures of Iron. John Lynch, esq., of Portland, Me.: Shipbuilding. J. D. Weeks, esq., of Pittsburgh: The Manufactures of Glass and of Coke. W. C. Wyckoff, esq., of New York: The Manufacture of Silk. Professor C. S. Sargent, appointed to report on Forestry, also takes the statistics of the Lumbering Industry. VII.-STATISTICS OF SPECIAL BRANCHES OF AGRICULTURE. In general, the statistics of agriculture can be collected only by enumerators of population in making the tour of their districts. Manufactures are usually concentrated in considerable villages or cities, so that a special canvass becomes comparatively easy and inexpensive. But it would involve a vast increase of the cost of the census were special officers, distinct from the enumerators of population, to be appointed to collect the statistics of the two and a half or three millions of farms in the United States. Nor does the same reason for a special canvass exist in the case of agriculture, as of manufactures, mining, or the fisheries. In agricultural districts the enumerator is likely to be a farmer, or at least farm-bred, and thus to know enough about such matters to be able to fill the schedules intelligently. At the same time, it has appeared to the Superintendent that the occurrence of the census affords an admirable opportunity for securing certain large classes of facts relating to land systems, labor systems, modes of culture, applications of machinery, etc., in agriculture, wholly in addition to the bare statistics of the crops produced, which, if justly collated, corollated, and illustrated, cannot fail to be of great interest and value. In this view, several special investigations of wide range have been undertaken, and others will be set on foot as a favorable occasion shall offer. The following is the list of experts and special agents appointed in this department of the Census: Professor E. W. Hilgard, University of California: Cotton Culture. Professor W. H. Brewer, New Haven, Connecticut: The Production of Cereals. Professor C. S. Sargent, Brookline, Massachusetts: Forestry. J. R. Dodge, esq., Washington, District of Columbia: Orchard Fruits, Tobacco, Hops. Clarence Gordon, esq., Newburg, New York: Meat Production in the grazing States and Territories. Professor Hilgard is assisted by a number of eminent agriculturists and geologists. Several states are being traversed for the purposes of the most thorough and exhaustive investigation of the conditions and methods of the cultivation of cotton yet undertaken by any government or association. Mr. Gordon has been in the field since July. The other branches of agricultural inquiry have more recently been set on foot. VIII-MORTUARY STATISTICS. But the chief effort made for the collection of statistics relating to the census year, in advance of the occurrence of the June enumeration, though not one requiring the appointment of an expert or special agent, has been in the direction of a mortuary record, to be kept by physicians and surgeons, of cases of death occurring in their practice. The United States are at a marked disadvantage, in comparison with almost any other civilized nation in the matter of vital statistics. We know not the number of persons born or dying in any year of our political history. The registration of births, marriages, and deaths, which in other countries is rigidly enforced by adequate provisions and sanctions of law, is in some states not even required by statute, while in only three or four of the states which maintain a formal registration is the service of such a character as to give any considerable value to the results. Mere provisions of law will not secure good vital statistics. There must be vigilant administration by expert and thoroughly trained officials, heavy penalties for delinquency, and a disposition of the public mind which will not only allow but demand the relentless enforcement of the law. It is only when it is popularly seen and appreciated that no one can be born into the community or die out of it without affecting the rights and interests of every preceding or surviving member that adequate legis lation and adequate administration will be provided for recording all the essential facts relating to the beginning and the close of every life. Outside the three, or at the most four states, above alluded to as maintaining a good system of registration, there are perhaps a score of cities which keep up something like a system of recording births and deaths, of which six or seven have established a reputation for the completeness and accuracy of their published reports. For all the rest of the country there is either no statistical information at all respecting the number of those who are born or die during any given period, or the statistics are palpably defective. The disadvantage to the United States arising from the lack of good vital statistics is most serious. Not to speak of the unenviable singularity which it gives our country among the civilized and progressive nations; not to speak of the uncertainty in which it involves our sanitary legislation and administration, or of the loss which the science and practice of medicine suffer from the absence of trustworthy information respecting the range and degree of virulence of certain fatal diseases and the rate of mortality in one section as compared with its neighbors; the mere pecuniary disadvantage at which our people are placed, in the important matter of life insurance, would, if truly estimated, far outweigh the whole first cost of good vital statistics. Hundreds of millions of dollars have been invested in life insurance in this country within the last thirty years, and yet we have not even an approximate life table (a) of the United States. Insurance companies do not know how much they should charge to be safe; the people do not know how little the companies should charge, to sell insurance at its fair value. All parties are and have been operating in the dark in the matter of interests involving enormous expenditures and receipts, for lack of information which only government can supply, and which in almost all other progressive countries government does supply. In partial recognition of the importance of mortuary statistics, the act of May 23, 1850, required the return, by the canvassers of population, of all deaths occurring a I speak with all respect of the effort made by Mr. E. B. Elliott to construct such a table for the Statistical Atlas of the United States in 1874. Mr. Elliott's effort was most praiseworthy, and his qualifications were equal to almost any task, but the fatal deficiencies in the information attainable during the census year; and if the provisions of the law had been adequate to its intent, the results would have been of great value, even though the facts were obtained but once in ten years. In truth, however, the statistics of mortality obtained through the census have always been defective, and often grossly misleading. In the Seventh Census, 1850, there were returned but 324,394 deaths from a living population of 23,191,876; in 1500, 394,153 deaths from a living population of 31,443,321; in 1870, but 492,263 from a living population of 38,558,371. It does not need to be said that such a ratio of deaths to living population is maintained in no considerable community of the world. Mr. Elliott estimates that in 1870 41 per cent. of the deaths occurring during the census year escaped record. The causes of such wholesale omissions in a periodical enumeration may be stated as follows: In some cases the canvassers fail to put the question; in others, heads of families, or persons mering for them, fail to recall the fact of a death occurring during the year, especially when ten or eleven months have already elapsed since the date of death, and the mind, not unnaturally, refers to the event as having taken place a year or longer before. In still another large number of cases persons be out of families, which class seems not to have been in contemplation of the census law, which makes the return of mortality a family return. In still other cases deaths occur in families, but the very death itself breaks up the family and scatters the surviving members, leaving no one to report the death in the census. In still other cases deaths occur in what are constructively families for the purposes of the census, i. e., boarding-houses, hotels, etc., but the common tie of membership or associ atom is here so casual and so slight that the chances are altogether against the circumstance being retained in memory six or eight months after. (a) In the provisions for obtaining vital statistics the act of 1879 differs from that of 1850, first, by allowing the registration of deaths, under state or municipal authority, to be substituted at the discretion of the Superintendent for the returns of enumerators; and, secondly, by placing it in the power of the Census Office to supplement the strictly official agencies by information derived from other sources. In view of the great importance of the subject, the earliest effort made after the organization of this office was in this direction. A small register was prepared sufficient to contain the record of twenty-four deaths, with a statement of the causes of death, the sex, age, occupation, and nationality of the deceased. A copy of this register was sent to every physician and surgeon, of whatever school, whose address could be obtained by the Census Office, with the request that the recipient would preserve therein a record of all deaths occurring in his practice during the census year. The most laborious correspondence was resorted to in order to form and perfect the list of physicians and surgeons for this purpose. Nearly 100,000 of these registers have been thus distributed. The response of the medical profession to this appeal has been most gratifying. Even while the list was being perfected, letters were received from hundreds of physicians offering co-operation, and asking to be furnished with a copy of the projected register. Almost without exception the medical journals and medical associations of the country, of all schools and names, have commended this effort, and urged their readers or members to contribute towards its success. The Superintendent cherishes the hope, which, he believes, is not beyond the reason of the case, that this scheme will secure a vast amount of information relative to the vital conditions of our country, which, under judicious treatment, will yield results of high scientific and practical value. THE COUNT OF THE PEOPLE. With respect to the house-to-house enumeration of the population, which is the primary and principal object in contemplation of the law, the act of March 3, 1879, makes a wide departure from the methods previously in use. Reports of Ninth Census, volume on Vital Statistics, pp. 192, 193, The more important changes may be briefly stated as follows: 1st. In the stead of adopting, as the units of supervision, districts (judicial districts) formed with reference to other and altogether different purposes, districts are to be formed wholly and simply with reference to the requirements and exigencies of enumeration. 2d. The number of such districts is increased to 150, more than double that of the judicial districts, securing a higher degree of local knowledge in preparation for the enumeration, and rendering possible a minuter supervision of the work while in progress. 3d. In the stead of imposing the duty of supervision in the districts, when formed, upon a class of officers (marshals of the United States courts) selected with reference to other and widely different services, and, in a large proportion of instances, crowded to the limits of their time and strength by prior official duties, officers (supervisors of census) are to be appointed solely with reference to their qualifications for the special and highly technical work of the census. 4th. The subdivision of the census districts for the purposes of actual enumeration is carried far below that required by the act of 1850, the maximum limit being now 4,000 inhabitants (according to the next preceding census), as against 20,000 formerly, and the Census Office being empowered to require a still minuter subdivision of the territory, if the good of the service shall seem to require it. By this change a high degree of local knowledge on the part of the actual canvassers is obtained as a security against defective and erroneous returns of the population. 5th. The period allowed for the enumeration is shortened to one month (the month of June) in rural districts and small towns, and to two weeks in cities, as against five months formerly. This change must result in obviating a large part of the errors resulting from the incessant movements of the population, especially during the later summer months. 6th. The appointments of the actual canvassers (enumerators) are made subject to the approval of the Census Office, a measure absolutely essential to good administration, but strangely overlooked in previous legislation. 7th. Instead of an inflexible rule for determining the compensation of canvassers, which did not recognize the difference in the labor of enumeration caused by the geographical features of the country, and which treated a square mile of river bottom or prairie as the exact equivalent of a square mile of rugged mountain, traversed only by broken roads or bridle-paths, the act of 1879 places the matter of compensation, so far as the canvassers are concerned, in the discretion of the department, which is thus able to combine the economy and efficiency of the service with justice to the persons engaged. 8th. The act of 1879 authorizes, at the discretion of the Superintendent, the use of "prior schedules", or blank forms distributed in advance, to be filled up with delib eration and after consultation between the members of a family; whereas, under the former system of enumeration, the canvasser, in the tour of his district, generally obtained the information from one member, not, as a rule, the head of the family, who was unadvised in advance of the subjects of inquiry, and was called upon to answer hastily a large number of questions relating to several different persons. The changes of system which have been noted vastly increase the work of the Census Office in making the preparation for the enumeration, but no one should be deemed fit for such a charge who did not rejoice in the added labor and care, in view of the manifold advantages to be obtained. THE CENSUS LAW. In making these preparations, through the six months that have elapsed since the organization of the service, the Superintendent has become more and more fully confirmed in the opinion that the legislation of the last Congress on the subject of the census was wise and salutary. Not a single fundamental defect in the scheme of enumeration has appeared, nor has any important change occurred to the Superin |