The census was taken for the night of Sunday, the 12th April, 1896. In the year 1891 the census night was the 5th April, so that the interval between the two enumerations of the people was five years and one week. The Act does not apply to Maoris, but a special census of the Native race was taken in February and March, under the supervision of the Justice Department. The following remarks (except when specially stated to the contrary) must therefore be understood as referring to a census of the population exclusive of Maoris, meaning by the term “Maori” any person of the aboriginal native race, and half-castes of that race living as members of any tribe.

The inquiries made of the people by means of the household schedule at the last census embraced the same subjects as before, with the exception that no information as to the land, cultivations, and live-stock was required, for the reason that these matters are being annually dealt with by the Department of Agriculture, Under the provisions of “The Agricultural and Pastoral Statistics Act, 1895.” Under section 9 of that measure the Registrar-General is required to adopt and use, for all purposes of –The Census Act, 1877,” the account taken by the Department of Agriculture, which is to supersede the account formerly taken by the Registrar-General in census years as part of the census itself. The relief afforded to Sub-Enumerators by means of the change of system was found to be very great.

The inquiry as to the domestic and social condition of the population was made under the following heads:—

The number and description of dwellings.

The names and surnames of the people.

Their relation to heads of households, as wife, son, visitor, lodger, servant, &c.

Their condition, as married or single, widower or window.

Their sex and age.

Their rank, profession, or usual occupation, present or past.

Their extraction, and place or birth.

Their religious denomination.

Whether British subjects or aliens.

Whether suffering from sickness, accident, or infirmity.

Whether able or unable to read and write.

If attending school (distinguishing kinds of schools, and between day- and Sunday-Schools).

Besides the above information as to all persons in the colony given in the household schedule, returns were obtained from proprietors or persons in charge of all manufactories, mills, works having machinery and plant, and other industries, including mines and quarries. The forms used for these were very comprehensive, and included questions as to hands and steam-power employed, value of materials operated on, and of produce or manufacture, with particulars as to quantities, an value of land, buildings, machinery and plant in use.

Further returns were collected under authority of the Fourth Schedule to the Census Act, showing the number of churches, chapels, and other places of worship, their situation, accommodation, and the attendance thereat; also as to land and building societies, mechanics’ institutes, with other literary or scientific institutions.

By section 3 of –The Representation Act, 1877—the Registrar-General is required to ascertain and report as early as possible to the Representation Commissioners the results of any census, and it then becomes the duty of the commission to divide the colony into electoral districts for the apportionment of the representation of the people on the basis of population. The final figures were accordingly formally reported in a certified return, on the 4th July, 1896. The return, which gave details of population to Parliament on the 7th July, 1896, following on a preliminary return made on the 11th June. Further census results were published in the form of summary tables in the New Zealand Gazette as under:—

Complete parts of the census volume were circulated as rapidly as they could be put through the Press, and issued on the following dates:—

31st August, 1896—

10th February, 1897,—

30th March, 1897,—

The cost per head of population of taking and compiling the census of 1896 is found to have been less than that incurred on the occasion of the census of 1891, on a comparison of the total amounts spent:—

As regards the expense of the enumeration of the people, excluding Maoris, the sum for 1896 is not indeed smaller than that for 1891, but a calculation per capita of population shows economy. The figures are:—

COLLECTION OF CENSUS (EXCLUSIVE OF MAORIS).

For the Maoris the expense was as under:—

COLLECTION OF CENSUS OF MAORI POPULATION.

For compilation and sundries in the central office, comparison shows a saving in 1896, though it must be borne in mind that land, cultivation, and live-stock returns had not to be dealt with the latter census. On the other hand, there is an amount of £257 included in the sundry expenses for 1896, which was spent on the enlargement and fitting-up of the special building used as a census office. No corresponding cost was incurred in 1891. The charges made by the Survey Department for mapping were far higher for the 1896 census than for that of 1891, the figures being £674, against £375 in the earlier year.

COST OF COMPILATION AND SUNDRIES (INCLUDING MAPS) IN CENTRAL OFFICE.

The total cost of the European census of 1896 was at the rate of 5˝3d. per head. In 1891 the total cost amounted to 5˝8d. per head. The difference on the side of economy is therefore one-halfpenny per head, notwithstanding the largely increased charge for maps, and the expenditure on building and office fittings.

A comparison of expenditure in the enumeration in proportion to dwellings is scarcely possible for different parts of the colony in regard to country districts, because distances between the houses vary, and for other reasons. But a comparison of cost per 100 houses in the four chief cities of the colony is both interesting and useful, as affording an idea of what the expense can be reduced to in centers like capital towns. The cost in 1896 was lowest in Wellington, £21s. 4d. per 100 dwellings. In Dunedin the expenditure was £3 4s. 4d. per hundred, and in Christchurch £3 7s. 4d,; while in Auckland the cost was as high as £5 4s. 6d., or considerably more than double that in Wellington.

The number of Enumerators appointed to control the taking of the census was thirty-two, against twenty-four in 1891. Each Enumerator had his particular district, consisting of a group of counties, with their interior boroughs. To increase the number of Enumerators' districts was essential, in view of the need for more supervision of the Sub-Enumerators employed to deliver and collect the household schedules than was exercised in 1891, and also for a more prompt examination and dispatch of the census to the head office by the local Enumerators. To allow time for the Commissioners to define the new electoral districts, and for objections to be raised to the proposed districts, for consideration of objections and making alterations, it was necessary to begin the final compilation at the earliest possible moment, which could not have been done without a larger number of Enumerators than were employed before. The number of sub-Enumerators was 787, against 708 in 1891.

The plan adopted previously, of forwarding to each Enumerator two large maps, showing the contiguous counties comprising his district, was again followed. One of these was sent back to the Registrar-General with the plan of proposed Sub-Enumerator’s districts marked thereon. The maps were drawn to show all the existing divisions of the country for purposes of general and local government, such as counties and ridings, road and town districts, boroughs, and towns.

The Enumerators were instructed to make a riding, or part thereof, a census sub-district, unless there were special reasons against taking such a course. Sometimes road districts were adopted. If the subdivision into Sub-Enumerator’s districts was not as minute everywhere as could be desired for purposes of the most rapid execution of the work, it was nevertheless done as closely in accordance with recognized methods as the circumstances of the colony would allow. It is sometimes found better to give nay thoroughly reliable person a rather large district than to subdivide further when suitable men are not available. But the subdivision of the colony as a whole was far more thorough at the census of 1896 than in 1891, which is shown by the relative number of Sub-Enumerators each of whom had a sub-district to himself.

Directly the plan of subdivision of an Enumerator’s district was settled the descriptions of the boundaries were forwarded to the Chief Surveyor’s Office, for maps to be prepared of each Sub-Enumerator’s d district, coloured to exhibit all territorial divisions. These maps enabled the Sub-Enumerators so to make up their books as to show the household schedules belonging to each division of the country.

The different Chief Surveyors, on completing the Sub-Enumerators’s maps for any particular county, transmitted them to the Surveyor-General’s office at Wellington, to be examined and passed on to the Registrar-General, who at once issued them (with books attached) to the Enumerator within whose district such county was situated. Later on the Registrar-General issued to all Enumerators full supplies of household schedules, with the forms for special industrial returns, and others relating to such matters as places of public worship, libraries, and other literary and scientific institutions.

After the census schedules had been all received from the Sub-Enumerators, a preliminary return of population was made up by every Enumerator, and from these a summary showing the rough results of the census was compiled in Wellington and presented to Parliament on the opening-day of the session. The first parcels of household schedules received from local Enumerators at the Registrar-General’s office came to hand on the 22nd April, 1896. The delivery went on from that time until the 11th June, 1896, when the last parcel arrived. In the meantime a large office had been opened in Wellington for the final compilation, and considerable progress made.*

The compilation of the census necessitated the employment of a large staff of temporary clerks. These were taken on gradually, as fast as the arrival of the schedules permitted. The maximum number of clerks employed was forty-two. The work of compilation began on the 25th April, 1896, and the staff was increased to the full number in August. Reductions began in October, and continued until, in February, 1897, there were only six clerks remaining.

These compiling clerks were selected from applicants in all parts of the colony.

Much trouble and delay having been experienced in 1891 for want of a proper office fitted to accommodate the compilation staff, and to hold the great mass of schedules, forms and other matters belonging to the work, a separate building, specially enlarged, was handed over to the Registrar-General in April, 1896, with the view of facilitating operations.

The census of 1896 was compiled in one room (40 ft. long by 16 ft. broad) so as to insure complete supervision of what was being done. By economizing space, clerks to the number of forty were seated in this room, besides the superintending officer. This arrangement was found to be a vast improvement on the previous plan, which was simply to borrow room for a few clerks wherever it could be found, and scatter the staff in small numbers about the Government Buildings.

A contrast to these complete arrangements for census-taking is afforded in the description of the method adopted in regard to Natives on the Gold Coast of Africa, given in the Governor’s letter to the Native kings of 22nd December, 1890, from which the following is extracted: “I understand the way you count your people is to divide each town or village into companies, which are again subdivided into families. The heads of families are then directed to drop into a calabash or similar article provided for the purpose a grain of corn, or cowrie, according to the number of their people—and that these calabashes are then collected and the contents counted. You will therefore have no difficulty in doing what I ask… When the numbering takes place, different articles are to be used for each sex—that is to say, Indian corn for males, and cowries or kernels for females.”

The first tabulation is made directly from the household schedules, and does not deal with any of the particulars relating to the people, such as age, religion, or birthplace. The number of each sex in every dwellinghouse is all that is noted, and the dwelling is classified as to number of rooms, and materials of which composed. The sheets on which these particulars are given are headed for every division of territory having boundaries existing in the colony. On completing the tabulation of the schedules belonging to any particular division of a county, a fresh sheet is taken, as the heading has changed. Thus, by breaking the sheets, summaries can afterwards be prepared without repeating any of the tabulation, no matter how frequently the boundaries of various divisions are found to interlace each other.

The schedules are laid out in counties, with what are termed “blue” and “white dividers” to mark where divisions, such as ridings, road districts, and localities, begin and end. The blue dividers simply indicate the localities or small places having names, but no legally defined boundaries.

The first compilation does not take very long to finish, and when done the Registrar-General is able to comply with the requirements of “The Representation Act, 1887,—by reporting the exact population to the Commissioners, who are appointed to divide the colony anew into electorates on the basis of the census, as before referred to.

An Act was passed on the 13th July, 1896, and two permanent Commissions of five members each were substituted for the one under the Act of 1887. One of these Commissions is for the North Island, and the other for the Middle and Stewart Islands. The official members of the North Island Commission are the Surveyor-General and the Commissioners of Crown Lands for Taranaki and Auckland. For the Middle Island the official element is represented by the Commissioners of Crown Lands for Westland, Canterbury, and Otago. The two remaining (unofficial) members of each Commissions first sit together as a joint Commission for the purpose of fixing according to the manner prescribed in “The Representation Act, 1887,” the number of the districts for the North and Middle Islands (including Stewart Island), respectively. Afterwards they act separately and independently.

The method laid down in “The Representation Act Amendment Act, 1889,” for computing for the purposes of that Act the population of the colony, is to add 28 percent. To the population not contained in any city, borough, or town district having a population of over two thousand persons. The total population of the colony (other than Maoris), with the addition aforesaid, having been ascertained, is then divided by the number of members (seventy), and the quotient thus obtained forms the quota. The four city electoral districts are so defined as to extent that the population shall be three times the quota. Inasmuch as it would be impossible to divide the country into a given number of districts all having exactly the given quota of population, the law permits the Commissioners to make and allowance of 750 persons by way of addition to or deduction form the population of rural districts, and 100 persons in the case of city electorates; and due consideration is given to community of interest, facilities of communication, and topographical features, as far as possible, in forming the districts. A table is given further on showing the new districts, with the actual population of each, and the nominal population—that is, with 28 percent. Added to the rural portion.

Proceeding as indicated above, the North Island was found to be entitled to thirty-four members, and the Middle Island to thirty-six. Previously the North Island had only thirty-one members, and the other thirty-nine, but movement of population has altered the proportions in favour of the North.

When the work of the Joint Commission was done, the Commissioners for each Island met in Wellington on the 27th July, 1896, to readjust boundaries, and again subsequently, on the 14th September (one at Wellington and one at Christchurch), for consideration of objections. The districts were finally gazetted on the 24th September, and the general election held on the 24th December, 1896.

Reverting to the compilation of the census, the second tabulation was not made directly from the census schedules. In order to obtain tables showing the number of people of each sex at various ages, in combination with information as to their religion, birth-place, conjugal condition, degree of education, occupation, health, and other particulars, it has been found best to mark a card for each person, and then to tabulate the results of the sorted cards on to sheets ruled with the necessary columns. Any number of different combinations can be obtained from cards with perfect facility. They are used in vital statistics for a double purpose, first as a means of tabulating causes of death in combination with ages of deceased persons, and then, after being sorted into an alphabetical arrangement of the names, the general index of deaths registered is prepared from them, by means of which searches are made when copies of entries are called for.

The labour of transferring information from the schedules to the cards in not nearly so great s might at first be imagined, because the cards are printed in such a way that a mere stroke of the pencil across a small square space is enough for the most part to indicate what is wanted. Recognised abbreviations of words are permitted for the rest, such as “G.S.” for Government scholar, “Dr.” for daughter. The accompanying specimen of the from of printed card used will show that much work is saved. A clerk in practice can mark a large number of cards in a day.

The checking of the cards is certainly a serious matter. It is effected by reading off with the schedules, and, if not insisted on, the census would be unreliable. But when the cards are all filled up ad checked the benefit is at once discovered, and certainty that the advantages far outweigh the delay and cost of making ready the cards. In fact, the work closes up rapidly when these cards are complete.

For further remarks on this particular from of card see paper on “Modes of Census-taking in the British Dominions,” by R. H. Hooker, M.A., Assistant-Secretary to the Royal Statistical Society, and read before that body on the 16th January, 1894. A model of the New Zealand card is there given, with special comments.

To avoid disturbing the cards for each county or borough by picking out those in regard to which exceptional information is wanted, a system of duplicate cards is used; as an instance, a second card is filled up for every Chinaman found in the census. These are kept by themselves, and tables relating to the Chinese are made up from them without interfering with the great mass of ordinary cards.

The processes of preparing the cards and checking them were carried on simultaneously, to secure good work. Card-writing began on the 10th June, 1896, and the whole were finished by the 4th September.

The sorting of the cards is done into cases placed upon the tables at which the clerks sit. These cases are divided into suitable compartments by means of movable pieces of wood, so that the sizes of the pigeon-holes can be varied according to the quantity of cards they are required to contain at different stages of the work. A checker tests the correctness of each sorting and count of the cards before they are removed from the pigeon-holes.

The Hollerith electrical machine, for purposes of dealing with the cards, is not suited for a country like New Zealand, where the population is not very large, and where the combination with the Canadian census of 1891:—

In tabulating the returns, the Hollerith electrical tabulating-machine (which should, perhaps, rather be called a totalling-machine), introduced in the United States, was used. For this purpose, a card, similar in principle to that used in New Zealand, is devoted to each person; but, instead of drawing lines, a hole is punched in the center of the compartment. Each card is then successively placed on a horizontal board. This board is pierced with holes corresponding in number to the total number of compartments on the card, and so situated that each hole is under the center of a compartment. Under each of these holes, again, is a tube partly filled with mercury, which communicates by means of a wire from the bottom of the tube with the index of a counter. Above the card is a second horizontal board, on the lower side of which are springs, terminating in blunted needles, these being so arranged as to dip into the tubes wherever there is a hole in the card, and thus complete an electric circuit wherever the needle meets the mercury. The electric current then moves the index of the counter through one division each time the board is lowered. By passing all the cards through the machine, the number of persons corresponding to each particular fact can be counted at once, and this number is then written on the tabulation-sheets. The machine is so arranged also as to particulars can be worked out by merely passing the cards through the machine. Two or three different combinations can be worked out simultaneously, provided that any one particular does not enter more than one of the combinations—e.g., the religion according to education, and the infirmities according to age, could be worked out at the same time. It is, I believe, recognized that the device would not have been of so much value in the United Kingdom and the other colonies, where the number of details required is not so great. Owing to the time occupied in punching the cards as compared with that of ticking the compartments, the economy only beings to be appreciable when the combinations are very numerous.

During the time that the second tabulation of the population was in progress the special industrial returns collected with the census were tabulated, and a complete set of tables relative to manufactories, works, &c., was compiled. Besides these, the returns of places of public worship, land and building societies, libraries, and other literary and scientific institutions, had also to be compiled. The complete census industrial statistics appeared in a Gazette of the 25th January, 1897; but tables relating to individual industries such as sawmilling, meat-freezing, butter- and cheese-making, had been in circulation long before that date. It was found that by publishing a little at a time the attention of the public was better drawn to the census, and interest in it kept alive, than by holding the matter back. After all the office summaries were completed, the tables for the census volume were made up from them. In these there is a good deal of calculating to be done, and many comparisons to make.

When the last of the summaries—those relating to occupations—had been completed from the cards, so that the entire census was represented in cards and sheets, permission from Government was obtained to burn the household schedules, about which there was trouble formerly. The householders give what is required by the Census Act on the condition that the contents of individual schedules shall not be made known to the public, but used by the Registrar-General for statistical purposes only. Therefore, these documents being confidential, it is undesirable to keep them longer than necessary, under any circumstances; but in this particular case the Registrar-General was called upon to give up the office-room used for the census, which was an additional reason for disposing of them. In Victoria the schedules are pulped, in the presence of an officer of the Statistical Department, directly after the compilation is finished.