THE COLONY OF TASMANIA, 1891.

PARTS I - VIII.

WITH INTRODUCTORY REPORT BY THE REGISTRAR - GENERAL.

TASMANIA :

WILLIAM GRAHAME, JUN., ACTING GOVERNMENT PRINTER, HOBART.

1893.

CONTENTS.

ERRATA ET CORRIGENDA.

INDEX TO REPORT.

REPORT AND APPENDICES.

INDEX TO TABLES.

PART I.—POPULATION, BUILDINGS, AND LAND.

PART II.—AGES OF THE PEOPLE.

PART III.—CONJUGAL CONDITION.

PART IV.—BIRTH-PLACES.

PART V.—EDUCATION.

PART VI.—SICKNESS AND INFIRMITY.

PART VII.—RELIGIONS.

PART VIII.—OCCUPATIONS.

ERRATA ET CORRIGENDA.

Page l., Report. Occupations of the People, after "diurnal tides," line 5, "full stop" should be printed, and next sentence start with "Kinds of service, &c."

Page lvi., seventh line from bottom of page,for"word's"read"world's."

Part I., Ages—

Page 7. In columns 2,3, 4, 5,fordecimal pointreada comma (,).

Page 18. Chinese 901,read931.

Page 45. Evandale, Females 698,read1576.

INDEX TO REPORT.

SECTION I.—INTRODUCTION.

SECTION II.—PROGRESS OF POPULATION.

SECTION III.—HABITATIONS OF THE PEOPLE.

SECTION IV.—AGES OF THE PEOPLE.

SECTION V.—CONJUGAL CONDITION OF THE PEOPLE.

SECTION VI.—BIRTH-PLACES OF THE PEOPLE.

SECTION VII.—EDUCATION OF THE PEOPLE.

SECTION VIII.—RELIGIONS OF THE PEOPLE.

SECTION IX.—SICKNESS AND INFIRMITY.

SECTION X.—OCCUPATIONS OF THE PEOPLE.

SECTION XI.—GENERAL PROGRESS SINCE 1881.

CENSUS OF TASMANIA, 1891. GENERAL REPORT.

TO THE HONORABLE THE CHIEF SECRETARY.

General Register Office, Hobart, 1st July, 1892.

SIR,

As the compilation of the results of the Census of Tasmania taken on the 5th April, 1891, has now been completed and published in separate Parts, it is customary and desirable that I should prepare an introductory Report for the purpose of giving a brief account of the events which led to the taking of the Census on 5th April, 1891 ; of the methods adopted in collecting and tabulating the information sought for ; and of the general results of the vast amount of information under various categories classified in the various Parts under the different Tables. I have endeavoured to carry out these objects in the following Report, which I have now the honor to lay before you.

SECTION I. HISTORY OF CENSUS-TAKING IN TASMANIA.

An account of the Eighth Census of Tasmania, taken on 5th April, 1891, simultaneously with nearly all other portions of the British Empire, would be incomplete without some brief reference to the history of previous Censuses.

Although official estimates of the population were prepared from time to time after the first settlement was established in the Island on 7th Sept. 1803,

The earlier Censuses were taken more frequently, the average interval being only about 5½ years. The last two Censuses (1881, 1891) were taken after longer interval ; the interval previous to Census of 1881 being 11 years, and the interval between 1881 and 1891 being 10 years ; thus—

• First Census 27th Sept. 1841.

• Second Census 31st Dec. 1847. Interval 6.26 years.

• Third Census 1st Mar. 1851. Interval 3.14 years.

• Fourth Census 31st Mar. 1857. Interval 6.08 years.

• Fifth Census 7th April, 1861. Interval 4.02 years.

• Sixth Census 7th Feb. 1870. Interval 8.84 years.

• Seventh Census 3rd April. 1881. Interval 11.15 years.

• Eighth Census 5th April, 1891. Interval 10.01 years.

Thus it appears the first proper Census of the Island was not undertaken until nearly 38 years after the date of its first settlement. This, however, is not to be wondered at when we consider that from the Middle Ages up to this time violent prejudices against the enumeration of the people prevailed in nearly all the countries of the world.

HISTORY OF CENSUS-TAKING IN OTHER COUNTRIES.

Undoubtedly it is due to such unreasoning prejudices that no attempt was made to take a proper Census of Great Britain until as late as the year 1753.

Happily the prejudices so strongly opposed to Census-taking up to the close of the last century have died away almost entirely, and there is now no civilized country which does not avail itself of the information which can alone be obtained by a Census, and which is so essential to the good guidance of all matters relating to the political, social, and material interests of the nation.

THE OBJECT AND UTILITY OF A MODERN CENSUS.

The object and utility of a modern Census are now so well understood by all thoughtful persons that it may perhaps be considered going beyond the purposes of an official Report to comment upon it. It cannot be overlooked, however, that among the people generally there still exists a large amount of misconception which operates seriously in marring the accuracy of results which can alone be obtained by theintelligent co-operation of every person in the community. I was profoundly impressed with the opinion that so long as such misconception obtains it is desirable to take every reasonable means to minimize it if it cannot be wholly removed. Accordingly, three months prior to the Census day a brief account of the objects, scope, and utility of a Census was drawn up by me in a popular form, for public circulation. With the approval of the Government, these observations were printed in pamphlet form, and, together with a circular inviting the co-operation of the Director of Education, the School Boards, School Teachers, and the local Press,—they were widely distributed throughout the Colony. By such means and by the hearty co-operation of the local press, the minds of the people generally were awakened and favourably interested.

The following extracts, bearing upon the utility, scope, and other important matters relating to census-taking, are taken from the pamphlet referred to, as they form an appropriate introduction to the analyses which follow :—

The Utility of a Census.

"The next great national "stock-taking," fixed for the night of 5th April, 1891, for all the members of the British Empire—comprising fully one-sixth of the population of the globe, and embracing at least one-fourth of its whole terrestrial surface—is looked forward to with the most eager interest by all thoughtful persons. Nor is this to be wondered at. The vital, material, and social well-being of a nation depends in a larger degree than is ordinarily realized upon the accuracy of its knowledge concerning the number, character, and condition of its people. It is the greatmeasuring-rodfor determining and making timely provision for the supply of all material satisfactions, and for guiding us more securely in determining growing tendencies—whether for good or evil—thereby enabling us to make accurate forecasts, and to make timely and intelligent provision for checking the evil and fostering the good. The objects of a modern Census are altogether different from those of the ancients. The Census of David, recorded in the 24th chapter of the Second Book of Samuel, was instigated purely by the spirit of aggressive warfare or conquest ; and, accordingly, its scope was confined to the number of "valiant men that drew the sword." The only warfare which the modern Census enumeration contemplates is confined to the peaceful conflicts of industry and commerce, and the discovery, subjugation, and control of the unknown or adverse forces of nature.

"Instead of the mere enumeration of valiant men who drew the sword, the modern Census seeks rather to ascertain the number of its "bread-winners," and the number of young, aged, and helplessdependants, in order to enable wiser provisions to be made for rendering the mutual exchange of services of the former more effective, and for making the training or care of the latter more in accord with the higher ideals of modern civilized life.

REPORT.

"But a Census enumeration does more than photograph the facts at the moment of Census-taking. It enables comparisons to be made with former periods, and from these we are enabled to determine many useful laws relating to the direction and rates of progression of all important movements and tendencies.

"Thusserious faminesin modern times, as compared with ancient times, are mainly averted, or their severity mitigated by statistical knowledge largely based upon the facts primarily ascertained by the Census. The lands devoted to agriculture and pastoral purposes year by year, if not wisely extended in harmony with the ascertained rate of the growth of population and its growing needs, would surely produce famine or misdirected and wasted labour, were it not that those whose interest it is to make provision are guided by statistical knowledge, which could only be rendered possible by periodic Census-taking. The termper headof the Statistician, though a meaningless symbol or phrase to the thoughtless or to the ignorant, is to the statesman, the life insurance actuary, the director of education, the commercialentrepreneur, the sanitary adviser, and social economist—analogous to the never-failing "barometer" or "sounding-lead"—whereby the course of events is instantly disclosed ; and thus making it possible to aid or counteract tendencies favourable or unfavourable, so far as such may be within human power.

"From what has been stated it is made sufficiently manifest that the great object of a Census is to gather such information as will be of material service in throwing light upon the more important social and economic questions, thereby making it possible to guide the affairs of the social organisation with greater intelligence and security as regards social and material well-being.

Scope of a Census.

"The value of information gained by means of a Census primarily depends upon the intelligence, honesty and disposition of the averagehouseholder. If we ask too many complicated questions we may put too great a strain upon average intelligence or comprehension, and the subsequent pretentions tabulation of the Statistician would be worthless or deceptive. If we arbitrarily seek to invade the privacy of the individual as regards means or income, morals, or even religious belief, we may certainly expect passive or indignant opposition, or—what is worse—we may receive, from many, dishonest replies. The practical Statistician who has had experience in Census-taking would avoid failure due to such causes by strictly limiting the number of enquiries to such subjects as will not unnecessarily invade the privacy of the individual, and not place a greater strain than can be safely borne by the intelligence and honesty of the pooresthouseholder.

"There is little or no value in the pretentions tabular results of countries where the scope of Census enquiries have been framed with too little regard for such important considerations.

The Weak Link of the Census.

"The strength of a chain is determined by itsweakest link,and the weakest link in Census-taking is thepoorly-educated householder.Many householders are unable to read or write, while a still larger number, from carelessness, ignorance, or lack of interest, fail to comprehend what is required of them in the filling up of the several columns of thehouseholder's schedule.The necessary notes of instruction at head of each column, amplified by more comprehensive notes of guidance and model forms of entry on the back of each schedule, appal and confuse many who are unaccustomed to interpretations of this nature.

"How can these difficulties be lessened, if not altogether overcome? Perhaps the most successful plan for securing and intensifying the interest of all persons in regard to the Census would be to secure the interests ofchildrenat school by the co-operation of School Boards, and especially by that of the localschoolmasterandschoolmistress.Let the Government call upon each Chief Superintendent of Census to prepare amodel lecturetteon the objects of the Census, and having done so, let there be issued a printed copy of the lecturette and a blank householder's schedule, accompanied by a courteously written circular addressed to thehead teachers of every school in the land, inviting their cordial co-operation by asking them to fix a special day for aCensus Lessonfor the children, accompanied by black-board illustrations as to the manner in which the householder's schedule should be filled up. If this were done it is certain that it would be productive of the very best results. The lesson of the children would certainly take strong hold of the interest of parents, and there is not a humble home throughout the land where the lesson of the children would fail to penetrate, and where the less favourably educated parents would not fail to become for the time being the proud pupils of their own more favourably educated children. The Press—the great educator—may do much to forward this desirable idea in the interests of the public weal.

Defects in previous Methods, and the desirability of greater Uniformity in different Countries.

"Although it has been inferred that greater firmness should be exercised in preventing enthusiasts from unwisely overloading the subjects of Census enquiry, still there is ample room for reform as regards the past methods of Census enquiry in the British Dominions ; more especially as regards defective classification ; their sterility in respect of important economic conditions ; and their lack of uniformity with each other. The Australasian Colonies hitherto have, in all essential features, followed the English Census method, and consequently reproduce its defects. Economists for many years have loudly proclaimed these defects, and their influence at the beginning of the year 1890 (February and March) secured the appointment of a Special Committee to enquire into certain questions connected with the taking of the Census in the United Kingdom.

"Prior to this enquiry in England, however, the representative Statisticians of Australasia had arrived at similar conclusions ; and in order that the next Census-taking in Australasia should be made as perfect as practicable, it was deemed advisable to obtain the consent of the various Australasian Governments to hold a Conference of Australasian Statisticians at some convenient place, for the purpose of discussing reforms, and for securing the general uniformity in methods of enquiry and in the final statistical tabulation of results.

CONFERENCE OF AUSTRALASIAN STATISTS WITH RESPECT TO THE CENSUS 1891.

"Their desires in this direction were finally realised by the initiative action of the Premier of Tasmania, (the Honorable P. O. Fysh), who issued a circular telegram to each of the Premiers of the various colonies of Australasia, as follows :—

Hobart, 23rd January, 1890.

To the Premiers of the Australasian Colonies.

This Government is desirous of arranging a Conference of Statisticians of the Australasian Colonies to decide upon uniformity of system for taking approaching Census. It is suggested that meeting should be in Hobart on Saturday, 1st March. Shall be glad to learn views of your Government by telegraph.

PREMIER,Hobart.

"The invitation, with one exception, was cordially responded to, and the following gentlemen, as representatives of their respective Colonies, met in conference at Hobart at or near the date appointed ; viz.—

Members of Conference.

H. H. Hayter, Esq., C.M.G., Government Statist, Victoria, unanimously elected by Members of Conference as its President.

R. M. Johnston, Esq., F.L.S., Registrar-General and Government Statistician, Tasmania.

T. A. Coghlan, Esq., A.M.I.C.E., Government Statistician, New South Wales.

H. J. Andrews, Esq., Under Secretary, South Australia.

E. J. von Dadelszen, Esq., Deputy Registrar-General, New Zealand.

Consulting Member—E. C. Nowell, Esq., (formerly Statist, Tasmania).

Secretary—Thos. C. Just, Esq.

"The Conference met on the 3rd March, in Hobart, and concluded its labours on the 18th March. The results of the Conference were deemed to be very satisfactory by all the Members, and the several matters which engaged their attention are summarized as follows :—

Extent and Nature of Subjects of Enquiry.

"The Members of the Conference unanimously deemed it unwise to extend the subjects of enquiry beyond the limits already determined in the Householder's Schedule of the Census for 1881, viz.—Name, Sex, Birth-place, Age, Religion (statement not compulsory), Occupation (stating whether Employer, employed on his or her own account, or Wage-earner), Conjugal Condition, Education, Sickness, and Infirmity, Materials of Houses, Number of Rooms. The model Schedule

Classification and Modes of Tabulation.

"As there were several imperfections arising out of defective classification and through lack of agreement in modes of Tabulation in the last Census, the Members of Conference were anxious to devise means for avoiding such difficulties or defects in the coming Census. At the same time they were impressed with the importance of adhering as far as possible to modes of classification hitherto employed, as without such cautious consideration useful comparisons with former periods and with other countries would be rendered impossible.

Tabulation.

"The methods for securing the best results by tabulation also engaged much attention. Theticket systemas now generally employed in the analyses of Vital Statistics was most favoured for speed, accuracy, and simplicity. Mechanical machines for counting the tickets speedily were also recommended for adoption, and these, with other plans for facilitating the work of tabulation, will no doubt have the effect of securing more accurate work and earlier publication of results.

General.

"The other matters, though of great importance to the tabulator and statistician, only affect matters of detail, and secure accuracy and uniformity. It is important to state, however, that one of the recommendations affecting the Census Act was, that only the heads of enquiry—not the Schedule—be embodied in the Census Act."

PREPARATION FOR TAKING THE CENSUS ON THE NIGHT OF 5 APRIL, 1891.

Preliminary work.

The success of a Census depends primarily upon the care taken by the Chief Superintendent of the Census, that the whole plan of operations has been thoroughly worked out in every detail at least nine months prior to the Census-day. This is absolutely necessary when we consider the vast amount of preliminary work which should be completed not less than two months prior to the actual distribution of Schedules throughout the whole country. Indeed in populous countries the time required for such preliminary work may engage the attention of a permanent staff during the whole of an intercensal period. The extensive character of this preliminary work, even in a small Colony like Tasmania, may be realised by a brief enumeration of the various forms, circulars, charts, and books of instruction devised, printed, and distributed in connection with the work of the last Census, in respect of a population of 146,667 persons scattered over an area of 26,215 square miles.

The principal of these are :—

1. Devising and printing schedules for distribution (1 to each dwelling). Printing, alone, equivalent to a 14 in. ribbon of printed matter 24 miles long, or 45,000 schedules.

2. Three sets of charts, scale 1 inch to the mile. A complete set for central office, one set divided among the Chief Enumerators of Census Districts, and one set divided among the various Sub-Enumerators of Census Sub-Districts. All these charts were carefully hand-coloured, showing the concatenations of each Electoral, Municipal, Registration, Road Trust and Census Sub-District, and each different division characterized by a distinctive number and boundary colour. This work involved the careful colouring of at least 100 square yards of charts.

3. Devising, classifying, and printing books of—

   1. Instructions for the special guidance of Enumerators of Districts. (See Appendix B.)

   2. Instructions for the special guidance of Sub-Enumerators. (See Appendix C.)

   3. Instructions for the special guidance of Tabulators. (See Appendix. D)

   4. Sub-Enumerator's compilation books.

4. Printing about 160 thousand tickets specially devised for facilitating the work of tabulation and ensuring greater accuracy.

5. Printing and distributing a large number of necessary circulars to Enumerators and Sub-Enumerators in connection with particular stages of the work.

6. Organising Census-staff, and carefully drilling them in their various duties, especially as regards the exact limits of their respective districts throughout the country.

7. Arousing the interest of the people and calling upon them to aid in making the work of the Census a success, by means of advertisements, pamphlets, and especially by enlisting the kindly co-operation of the Press.

The foregoing enumeration suffices to show that a large amount of important work is involved in the preliminary requirements of Census-taking, and necessitates—long prior to the Census day—the careful study of countless minutiae, the oversight of any one of which might destroy the value of the whole work of the Census.

The work of Enumeration.

The work of Enumeration, and the methods and precautions taken to secure completeness and accuracy, are usually mixed up in the larger number of descriptive reports on such matters. In this Report I have in a large measure eliminated the minutiae which refer to the methods and precautions taken to secure completeness and accuracy—which can be better studied apart in appendices—in order to impress more clearly upon the mind the extent and character of the actual work of Enumeration.

The area to be covered.—

In the first place it must be borne in mind that Enumeratorspersonallymust visit,twice

1. The density of population.

2. The distance which one person could cover per day either on foot or on horseback.

The greater the density of population enables the Collector to cover a larger population in a given time, but the work involved lessens his power to cover a large distance in traveling ; and conversely, the smaller the density of population enables the Collector to cover a larger distance in travelling. Hence it is manifest that in mapping out the work of Sub-Enumerators the size of their respective Sub-districts should in a measure be determined inversely in relation to the greater or lesser density of the population.

By estimates based upon such considerations, aided by the local topographical knowledge of the Chief Enumerator of each Electoral District in the Colony, it was determined that the

26,207 square miles of Rural Districts could be covered by 256 Sub-Enumerators ; thus, on the average giving 102.35 square miles, 85.75 dwellings, and 408 persons to each Sub-Enumerator. The Urban Districts (Hobart and Launceston) were likewise determined, and arrangements made provided that the more densely populated urban area of 7.36 square miles could be so divided that the work might be efficiently carried out by 39 Sub-Enumerators, thus giving an average of 0.19 square miles, 201.2 dwellings, and 1078 persons to each Sub-Enumerator.

The following summary shows in all these respects the apportionment of work of Enumerators for the principal divisions of the Island forming the larger Census Districts, which for the most part coincide with Electoral Districts :—

Thus, upon the whole, the actual Enumeration was carried out successfully by 295 persons, being 1 to every 88.86 square miles ; 1 to every 101.4 dwellings ; or 1 to every 497 of the population.

Cost of Enumeration.

The Cost of Enumeration is mainly determined by the work involved (1) in traveling from house to house, and (2) the number of the population entered upon the house-schedule.

In sparsely populated districts the greater proportion of the time and expense of the Enumerator is absorbed in traveling from dwelling to dwelling, the proportion in each district varying with the density of population—increasing relatively with sparsity, and decreasing relatively with density. The relative costper head,therefore, progresses inversely to absolute cost, and to relative costper square mile.Hence it follows that as a population increases in density the cost per head becomes less and less; while theabsolute cost,andcost per square milereally increases with the density, although at a differently proportionate rate. Thus it happens that the densely populated Urban Districts of Tasmania (Hobart and Launceston) in 1891 only show a cost of 0.78d. per head for Enumeration, while the Rural Districts cost 3.86d. per head. Reckoning by area, however, reverses this order—the Rural cost of Enumeration being only 1s. 4d. per square mile covered by Enumerator, while the cost in Urban Districts is as high as £18 9s. 7d. per square mile. The cost of the Enumeration of a progressive place, district, or

country should show an absolute increase at each succeeding Census, and relatively the cost per square mile should progress inversely to cost per head, the former with an increasing tendency, the latter with a decreasing one. The following is a comparative statement of the absolute and relative cost of Enumeration in Tasmania for the last three Censuses :—

The same law as to cost of enumeration is also revealed between Countries differing in density of population ; and between themselves at different Census periods, as shown in the following illustration :—

The study of the above table shows clearly how costper headin all countriesdiminisheswith increasing density of population, while a contrary increase in absolute cost and costper square mileinvariably happens with increase in the density of the population. A low costper head,therefore, taken by itself is no test of economy ; for it can only be reduced as theabsolute costandcost per unit of areaincreases. The total cost of enumeration for Tasmania in 1891 being only 1s. 5d. per square mile, and 2.98d. per head, compares very favourably with former periods in Tasmania, and, on the whole, compares most favourably with the cost of enumeration in other countries when both factors are taken into consideration. Indeed, the consideration that the absolute cost of enumeration is actually less by £76 than it was in the Census of 1881, while the population enumerated in 1891 is 26.73 per cent. greater, is the best evidence that can be afforded of the care taken to secure the information required at the lowest cost consistent with efficiency.

The Systematic Work of Tabulation, Summation, and Analysis.

The systematic work of tabulation, summation, analysis, and check involved in a modern Census is enormous. This work is wholly carried out in the central Office after the Schedules and Summaries of Enumerators have been gathered, checked, and arranged in classified order. Primarily, the extent of this work is determined by four fundamental considerations :—

1. The number of the Population.

2. The number of distinct categories under which information regarding the same person is again and again abstracted, re-arranged, computed, and set down.

3. The number of distinct subdivisions under each category for which summaries are prepared.

4. The large number of calculations involved in determining the relative per-centage proportion of the figures contained in the numerous summaries under the various categories.

The information requiring to be compiled in the Census of 1891, according to the plan agreed upon by the Conference of Australasian Statisticians held in Hobart in 1890, embraces a larger number of categories than has ever been attempted before in the Census of any country.

This is best shown by comparing the categories of Census Schedule of Tasmania for 1891 (excepting Dwellings and Land) with those of the Census of 1881, and by estimating the comparative number of distinct groupings into which the population must be arranged under the various categories in the original compilation sheets alone, to obtain the necessary information afterwards summarised :—

The number of separate groupings of individuals under the various categories necessary to arrive at specific aggregates, afterwards arranged in classified summaries, is greater on the whole by about 140 per cent, than in the previous Census of 1881. Indeed, the specific elaboration has been multiplied about 3.28 times ; but a great deal of saving was effected by not repeating the minor points of analysis under Municipal and Registration District tables (as in the Census of 1881), as it was deemed to be sufficient to confine such analysis to one system of geographical subdivision—the Electoral. The total cost of enumeration, tabulation, furniture, stationery, and instruments, amounted to £3380 ; that is, £365, or 12.10 per cent., above the actual cost of the Census of 1881.

Seeing that the work of enumeration embraced a population 26.73 greater ; and that the work of tabulation has been increased fully three-fold, the increase of only 12.10 per cent, in cost is most satisfactory. This result would not have been secured had it not been for the improved method introduced by me for facilitating the work of tabulation, based upon theticket system(hereafter described), hitherto applied successfully in the Registry Department in the preparation of the monthly and yearly analyses of Vital Statistics. A wonderful saving of labour in the many thousand calculations of proportions was also effected by the use of Tate'simproved Arithmometer,

Method of Tabulation.

To obtain with accuracy the aggregates under the subdivisions of each category, from the general information relating to 146,667 persons, contained in 29,801 schedules, may seem a small

matter when contrasted with the tabulation of the results of a large population, such as the United Kingdom or the United States. The labour of compilation, however, is not fairly ganged by the differences in population. It is true that the preliminary work of abstracting the particulars—provided the categories are the same—is fairly measurable by the number of the population ; but this need only absorb aboutone-fifthof the labour of permutation and summation, which latter may depend far less upon the number of the population than upon the number of the groupings for which aggregates have to be obtained and summarised under the various categories. Take, for example, the earliest, Censuses of which we have any account, viz., the Census of Moses recorded in Numbers (i. l-46), and Census of Joab (1 Chron., xxi. 1-17 ; 2 Sam., xxiv. l-17), of which the following is an abstract :—

Neither of these Censuses embrace more than about 21.62 per cent. of the population, and contain only one simple category, viz., "valiant men who drew the sword ;" and yet the simple enumeration of the later of these Censuses took 9 months and 20 days to accomplish. Compare this with the enumeration of the 62,000,000 persons of the United States, 1890, accomplished within a month, and its cities of over 10,000 inhabitants within a fortnight. Compare also the Census enumeration of Tasmania, an area nearly four times the size of Ancient Palestine, which was accomplished within ten days of the night fixed for the Census, and the results giving the numbers and proportions of both sexes, for every Electoral District in the Colony, together with elaborate comparisons with former Censuses, were actually published on 27th April, 1891, that is within 22 days of the time actually fixed for taking the Census, About three months later the results of tabulation were so far advanced as to enable me to publish the more important particulars relating to seven out of the nine principal categories, viz., Population ; Dwellings ; Ages ; Education ; Conjugal Condition, Birth-places, Religions. With these were contrasted corresponding points of information relating to the seven previous Censuses so far as such information was obtainable. At the same time a classified statement of the distribution of the people was given for each Electoral and Registration District, together with information in each case relating to area and density of population. These advance Reports satisfy the more important needs, and compensate in no small degree for the necessary delay in getting the completed voluminous tables of details passed through the press.

Method of Tabulation by the Ticket System.

Although the 200,000 or more groupings under the various categories relating to 146,667 persons in the Census tabulation are enormous, they are simply the result of the necessary permutations of 11 simple categories ; viz.—

A ticket was thus prepared for each one of the 146,667 persons living in Tasmania on the Census night, and the persons employed in writing up tickets from schedules were paid for the preliminary work at the rate of 2s. 6d. per 100. The tickets were conveniently tied up in 2-set packets, males and females, one 2-set packet for each of the Census subdivisions, corresponding to the unit of any of the district divisions for which results had to be obtained. Tied with each packet was a small label or tab-ticket, upon the which was carefully marked the total number of tickets in packet in conjunction with the number of Census subdivisions. These packets were carefully locked away in cell-presses according to a classified system of Census subdivision localities, so that the person in charge could deliver or replace in its classified position any packet at a moment's notice. Thus all danger of loss or disorder was obviated.

To give the utmost facility to the appointed tabulators special simple tabulating tressle-tables 8 ft. by 3 ft. 6 in. each were provided, each furnished with a lock-drawer underneath, and a vertical frame of ticket-cells fitted on to the back to secure the groups at various stages of the process of tabulating ; two ink-bottles, one for red ink, and one for black ink, were sunk flush with surface at each near corner ; while the surface of the table was divided vertically and horizontally by clearly marked lines, for arranging the tickets in distinct heaps in accordance with the particular classification in process of being carried out.

Much thought was necessary to secure the greatest amount of work by the least amount of labour ; but a few simple rules were devised and printed for this object, and also for preserving uniformity of procedure. The great principle, however, in beginning the work of tabulation relating to any stage was to fix upon thecategory and its subdivisions which were in every case common to several others—a constant, so to speak—and having so determined the category for the initial stage, then to break up the packet of the particular Census subdivision, and carefully arrange the cards at the top or side, as the case may demand, into as many subdivisions as, by previous classification, might be determined. Having distributed them in classified order under this first category, next to count the number of cards composing each subdivision, at the same time recording the total of each group in the corresponding subdivision of this category in tabulating sheet. Finally, adding all such divisions, to prove that the aggregate agreed with the total number of tickets in the Census subdivision packet. This base of constants for other categories being proved, the tabulator would commence the next category by breaking up and dealing in the same way with each of the subdivisions in succession, of the first category or

constant, recording and checking with ascertained total as before. When second category was completed, he carefully restored each group of second category to its place, as a constant, in the original subdivision of the first category, and proceeded with the third, fourth, or more categories until all were complete in relation to the constant or first category selected. The small size, smooth surface, and stiffness of the tickets favoured rapidity in sorting and counting in orderly heaps on table along the marked divisions,in the smallest possiblespace ; the latter consideration a matter of no small importance.

The constants common to several other categories—not reckoning sex or locality constant, which of course is preserved by using distinct tabulating sheets—were :—

1. Age in its fullest subdivision (embracing Conjugal Condition, Education, Instruction, Sickness).

2. Religions in subdivisions of classified Denominals (embracing Education, special groups of Age).

3. Birthplaces in subdivision of Country-born (embracing Education, Special Age groups, special grouping under conjugal condition).

4. Occupations, in subdivisions of specific Occupation (embracing Special Age groups, Grades, Districts.)

By such system the whole of the categories and their varied subdivisions were tabulated and checked in a very satisfactory and orderly manner. All unnecessary handling was obviated, and the greatest possible results were accurately secured with the smallest possible expenditure of labour.

The compilation of the Census subdivision results, under all predetermined categories, with their subdivisions were afterwards aggregated to form results under the various Districts under which it was determined they should be printed, whether in the detailed or summarized forms.

Where the number of cards under ultimate subdivision are very numerous, as in populous countries, the use of automatic ticket-counting machines would be required to facilitate labour ; but in counting the ultimate subdivision groups of a small population like Tasmania, the initial expense of such automatic aids would, upon the whole, cancel the advantage saved in hand counting.

My attention at the first was also directed to an ingenious machine employed to some extent in America and elsewhere, whereby the counting under all categories is effected simultaneously by electrical "make-and-break" contacts. But, so far as I was able to learn, the advantages are not so great as might be supposed, as every distinct particular under each category must previously be indicated by holes punched by hand. The abbreviated signs on cards, I believe, can be done far more speedily.

SECTION II. PROGRESS OF POPULATION.

The results of the Census Returns, 1891, as in the following summary, show that, though the population has made a greater absolute increase than in the interval of any previous Census period in Tasmania (viz., 30,962 persons, or 26.76 per cent. over Census of 1881), and a greater relative rate of increase per annum (2.40 per cent.) than in any Census interval since the year 1861, yet it falls short of the estimated population for the year 1891 (1st January) by about 10,000 persons, or 6.88 per cent. This error of excess in the estimates—based upon the balances of Births and Deaths, and Migration Returns—is undoubtedly due to an accumulation during the ten years of a constant error in the Customs Returns of Emigration. In the light of the present Census Returns it is estimated that the figures relating to emigration involve a constant error of understatement amounting to about 7 per cent.

This error of understatement as regards emigration returns is a general one in all countries ; but, during the two years, 1889-90, it is affirmed that a very large number of excursion travellers taking return tickets in other Colonies have wonderfully increased, and it is now inferred that no account of these have been taken in the return journey from Tasmanian ports, and this omission has greatly intensified the usual amount of error of understatement. It will be necessary henceforth to qualify the actual returns of migration to prevent a corresponding over-statement in all future yearly estimates of population.

The above analyses convey much information regarding the progress of population at each Census year in relation to the immediately preceding Census, as well as to the period 1841 to 1891. The particular causes affecting the increase or decrease of population at each period reveal how much is due to natural increase and how much to the balance of migration. The latter has been against the Colony's progress in the intercensal periods 1847-51, 1861-70, and 1870-81 ; but over the whole period 1841-91 it contributed 20,886 out of the total net increase of 96,451 persons, or 21.65 per cent. The natural increase in the intercensal period 1881-1891 25,001, is the highest on record absolutely, and the true rate per cent. per year (1.97) has only been exceeded by the intercensal period 1857-61 when it had been abnormally high, viz., 2.00 per cent. per year. The mean rate of natural increase per cent. per year since 1841, is equal to 1.87 per cent. This high rate of natural increase over such a long period is the best proof that can be given of the healthful condition of the Colony, being 110 per cent. higher than the natural rate of increase of London during the last thirty years, and 36.50 per cent. higher than the average natural rate of increase in England and Wales for the same period.

NOTE.—The boundaries and number of Districts were much altered between the two last Censuses, and hence it is difficult to make strict comparisons between many of the Districts. Taken in the groups as bracketed, and by the larger divisions, the comparisons are approximately correct.

The order of the various Electoral Districts showing the greatest relative increase and the least relative decrease since 1881 is shown as follows, beginning with the highest relative increase :—

From the above summary it will be seen that about a fifth of the Districts show a slight decrease as compared with the former Census. The greater number, however, show increases ranging from 8.46 per cent. (Franklin), to 132.15 per cent. (Cumberland, including the Mining District, Macquarie).

The general movement of population, however, is best appreciated when studied in relation to the larger geographical divisions, as in the following summary :—

That is, one and two are above the average increase of the Colony as a whole, while three and four are below it.

The order of the greater absolute increase is different however ; thus :—

Relative order of Importance.

The proportion of population in each division in relation to the whole Colony, although slightly modified since 1881, does not distrub the order of their importance as ascertained in the previous Census ; thus :—

COMPARATIVE PROGRESS OF THE POPULATION OF CITIES, TOWNS, VILLAGES, AND MINING CENTRES IN THE LAST INTER-CENSAL PERIOD.

The following abstract has been specially prepared to indicate the comparative rate of growth of population in cities and various small centres of population which had been in existence prior to the Census of 1891 ; and also to show, in a general way, the trend of what may be termed "the drift of population" as in contradistinction to "growth by natural increase." The artificial limit of comparison to centres over 100 persons, however, conceals the actual or true rate of growth in the case of those small villages or communities which during the last intercensal period have passed from below to above the limit which the necessities of tabulation restrict recognition as distinct centres as considered apart from the population sparsely scattered throughout the country in a less gregarious state.

From the foregoing analysis we learn that the cities, towns, and villages, &c. over 100 inhabitants, which numbered 43 in 1881, had increased to 79 in 1891 ; i.e., an increase of 26 ; and the population of all such centres had apparently increased by 26,899 persons, or 48.95 per cent. This, however, must be qualified by the consideration that the population of the 26 new centres in 1891, contain also the unknown element under 100 persons which existed in a number of localities when the previous Census (1881) was taken. With this qualification, however, it is clear, on the whole, that the city, town, and village populations have increased at a higher ratio than the residue scattered throughout the Colony engaged in the primary industries of farming, mining, or timber-getting. The highest ratio of increase, 239.3 per cent., is to be found in the North-Western Division, and the lowest, 29.56, in the South-Western. The greatest absolute increase (8926), however, is to be found in the North-Eastern Division.

The following centres show rates of increase since previous Census above or below the average of the whole Colony (26.76 per cent.) as indicated in descending order :—

Centres showing a Decrease.

The following nine Centres show a positive decrease as compared with the previous Census. It is significant, however, that none of them is included in the more progressive region of the North-west :—

New Centres having more than 100 Inhabitants.

The following is a list in the order of their importance of the 36 new Centres of over 100 inhabitants which have grown up or have become established since previous Census :—

Ulverstone, Dundas, Invermay, Wellington Hamlets, Glebe Town, Bellerive, Ellesmere, Glenorchy, Strahan, Mathinna, Penguin, Derby, St. Mary's, Trevallyn, St. Leonards, Mangana, Cressy, Gladstone, Cornwall, Weldborough, Woodbridge, Winkleigh, Margate, Low Head, Huonville, Bridgewater (North), Ramsgate, Breadalbane, Leith, Ringarooma, Forth, Hagley, Branxholm, Bracknell, Mount Nicholas.

Summary.

Thus, in summarising the results between the two Censuses, we find that out of the 79 enumerated centres in 1891—

SECTION III. HABITATIONS OF THE PEOPLE.

The habitations of all descriptions returned on the day of the last Census numbered 29,801, being an increase of 6272, or 26.65 per cent., as compared with the number of habitations returned in the Census of 1881. The dwellings per square mile number 1.07, and the persons per dwelling, 5.23.

General Character of Buildings, &c.

Of the 29,801 habitations returned in 1891, the following is a summary showing the general character of the buildings, &c. used as dwellings :—

HABITATIONS OF THE PEOPLE AS CONTRASTED WITH FORMER PERIODS.

Although from the previous summaries we may readily analyse particulars relating to the occupancy and character of the dwellings of the people, it is only by contrast with former periods or other countries that they become significant to us in determining the changes that may have taken place beneficial or otherwise.

The following comparative table of the four Censuses since 1861 is designed to show that under the principal categories the changes that have taken place in this interval, so far as information can be ascertained :—

From the preceding table we perceive that in nearly every respect there has been a great improvement in the condition of the people, as indicated by the habitations in which they live.

The increase in inhabited houses during the last decade—6616, or 28.21 per cent.,—is much higher both absolutely and relatively than during the three preceding intercensal periods. The number of dwellings per square mile in 1861 was 0.67 ; in 1870, 0.69 ; in 1881, 0.83 ; and in 1891 the number increased to 1.07 per square mile. Notwithstanding the considerably larger number of persons, mainly adult males, living in tents in 1891, caused by the larger amount of work in outlying districts carried on by the Government in opening up new districts by the construction of roads and railways, the proportion of persons living in one and two-roomed houses is considerably less than in the previous Census. Thus, the percentage proportion of persons living in houses of three rooms and over in the year 1881 was 72.65 per cent. In 1891 the percentage rose to 77.75 per cent. This indicates a vast improvement in the homes of the people, and represents a very large percentage when compared with the manner in which the

masses of the people are lodged in the great cities of the United Kingdom. In Glasgow, for example, only 30.01 per cent. of the people are lodged in houses of three rooms and over. This conclusion is also corroborated by the proportion of persons living in houses whose yearly rental is £20 and over, in Tasmania and Great Britain respectively ; the proportion in Tasmania being 34.49 per cent. of the population, whereas in Great Britain it is as low as 15.34 per cent. A similar contrast could be furnished in favour of the neighbouring Australasian Colonies. No better index of the improved condition of the mass of the people of these Colonies could be given than that afforded by the above contrast, wherein it is shown that about 20 per cent. of the lower industrial class, by improved local conditions, have been raised into the class next above them in order in Europe.

The distribution and density of dwellings in the various Electoral Districts are given in the following table :—

SECTION IV. AGES OF THE PEOPLE.

To ascertain the exact number of persons living at each age-period, is one of the most important objects of a Census enquiry. The knowledge which has enabled the higher civilized communities to accurately gauge and provide for the better education of the young ; the maintenance and improvement of the health of communities ; the beneficent provision for the support of the widow and orphan ; the care of the young and helpless in factories ; and many other matters relating to the welfare of the people, depend most intimately upon the accuracy with which we gauge the number of the living and dying at each age of life. Owing to various causes it has always been found impossible to obtain by Census method of enquiry a strictly accurate statement of the ages of the people at each single year of life. The reason for this defect has been clearly stated by my friend and predecessor, Mr. Nowell, in his valuable Report upon the Census of 1881. At page xviii. Mr. Nowell observes that "many people utterly fail to realize the fact, that it is the absolute duty of every loyal citizen to afford the most correct information which it is in his power to give, when the same is required of him by the State......... and a very considerable number of persons, not only in the uneducated, but even in the educated classes, really do not know their own age, —and there are many parents who have forgotten the (exact) ages of their children." Here, as in all other countries, there is a tendency among such people to put themselves down at the nearest round period, notably at 30, 40, 50, 60, 70, and 80. These round decennial periods act like attracting cores round which the unconscientious and those whose ages are imperfectly known segregate and cluster at the expense of the one or two contiguous years ending in the lower 9, or in the higher 1, and especially the latter. The constancy of these false nodes at the decennial periods is shown graphically on the opposite page, and may be illustrated by the following figures taken from the detailed statement for 1891, (Table l.) :—

Although the accidental intrusion of fresh immigrants may account for some proportion of the irregularities from the law of decrease in each successive year, it is evident from the intensity and regularity of the segregation at the decennial nodes that they are mainly due to the causes particularly referred to. The same law of segregation is also discernible at many of the higher quinquennial periods, but in a more subdued form. Directly, therefore, the number of each single age cannot be obtained from the figures actually given ; indirectly, however, the statistician can arrive at a very close approximation of the truth by determining the aggregates of each ten years based upon the more accurate quinquennial periods, and afterwards distributing them proportionally into single years by the average law of yearly decrease. In practice this method has been found to approximate very closely to the truth, and to afford a practically safe foundation in the hands of a skilful actuary for the safe conduct of all important matters relating to annuities and life assurance. For other useful objects the information gained by the ordinary quinquennial periods fulfils all purposes with the utmost perfection.

With these preliminary observations regarding ages, we may now proceed to single out the more important generalizations for consideration and comparison.

Of the 146,677 persons returned at the Census of 1891, the ages of 146,077—76,990 males and 69,087 females—were specifically recorded in the various schedules, of which the following is a correct summary by quinquennial and other useful groupings :—

The above summary gives in a convenient form all the most useful generalizations relating to age. To make it still more serviceable for general information the whole of the population, including the 590 unspecified, are also grouped approximately under headings, which convey in a more natural way* the composition of the people as determined by age and sex, as in the following summary :—

Proportion of Persons living under principal Age-groups contrasted with similar Groups at the Three preceding Census Periods.

The following summary is designed to show how far the proportion of persons living under important age-groups in 1891 agree with or differ from similar groups at the three previous Censuses, 1861, 1870, and 1881 :—

The proportion of persons living under 5 years and over 65 years is singled out, as the proportion of children and aged people in a community largely determines the importance of averages such as Birth-rates and Deaths, which are usually calculated from the total population, without necessary qualifications as to the age-composition, upon which its value rests for comparative purposes. In the first three groups, 0-5, 5-65, and 65 and over, the proportion of each for 1891 differs only in a slight degree from corresponding groups in 1874 and 1881. The number of males in 1891, at the most vigorous period of life (20 to 40), termed thesoldier's age, however, shows that the effective strength of the population is much higher than at any of the previous Censuses. This is accounted for by the larger proportion of male immigrants arriving in the Colony during the last decade, attracted thereto by the unusual impetus given to trade and industry by the construction of roads, railways, and other public works carried on by the Government in opening up the public lands of the Colony.

Proportion of the Population living in Tasmania in 1891 within principal Age-groups, contrasted with corresponding Groups in other Australasian Colonies.

The full significance of the proportions at particular age-groups can only be appreciated by comparison with other countries. But even then caution must be used in drawing inferences from particular cases. A relatively low or high proportion of any age-group may be the effect of very different causes. Thus, the proportion of the effective population—ages 20 to 65—may be high from two widely different causes. In Queensland, the high proportion of group 20-65—viz., 52.01 per cent.—is mainly the result of theintrusion of an abnormally large number of immigrantsfrom other countries. In France a much higher proportion is maintained within the same age-group, owing to theabnormally low birth-ratetending to produce an almost stationary population. Another cause tending to increase the older age-groups may be found in countries where, owing to climatic causes, the death-rate of children under 5 years is excessively high. Thus, immigration,—a stationary population caused by a low rate of fertility,—and a prevailing high death-rate among children, may equally be potent in producing an abnormally high proportion at the higher age-groups. On the other hand, a high rate of fertility tends to diminish the proportion of ages at all the higher age-groups ; and this, together with the fact that there is realatively a smaller intrusion of immigrants among the population in Tasmania, sufficiently accounts for its somewhat smaller proportion of persons at the worker's age—say 20-65—(viz., 46.65 per cent.), and its larger proportion of non-effectives at ages 0-20 and 65 and over—viz., 53.35 per cent.

The following table conveys in a convenient form the proportions of the people of Tasmania at the more important age-groups, as contrasted with similar groups in the neighbouring Colonies as ascertained by the last Census :—

Proportion of various Age-groups in Tasmania contrasted with similar groups in various Countries.

The two following tables have been prepared to illustrate more particularly the effects upon age-group proportions arising from differing causes as indicated in previous remarks :—

France affords the best example of a comparative stationary population ; and it is interesting to observe how closely its proportions at the various age-groups correspond with those of an Ideal Stationary Population as determined by George King (Text Book, Institute of Actuaries, 1887.)

We may now with advantage summarise the composition and order of the seven Colonies according to the relative proportion of the respective populations living within the principal age-periods, beginning in each with the highest, thus—

Relative Percentage Proportions in descending order.

Infants and Children, 0-5years.—(1) Queensland, 15.41 ; (2) Tasmania, 14.69 ; (3) New South Wales, 14.68 ; (4) South Australia, 14.13 ; (5)Australasia, 14.00 ; (6) Western Australia, 13.78 ; (7) New Zealand, 13.30 ; (8) Victoria, 13.10.

Boys and Girls, 5-15years.—(1) New Zealand, 26.72 ; (2) South Australia, 25.25 ; (3) Tasmania, 24.55 ; (4) New South Wales, 23.70 ; (5)Australasia, 23.49 ; (6) Queensland, 22.03 ; (7) Victoria, 21.52 ; (8) Western Australia, 20.55.

Youths and Maidens, 15-20years.—(1) New Zealand, 10.42 ; (2) Victoria, 10.08 ; (3)Australasia, 9.82 ; (4) South Australia, 9.67 ; (5) New South Wales, 9.67 ; (6) Tasmania, 9.51 ; (7) Queensland, 9.03 ; (8) Western Australia, 8.68.

Young Men and Women, 20-30years.—(1) Western Australia, 22.59 ; (2) Victoria, 21.45 ; (3) Queensland, 21.36 ; (4)Australasia, 19.63 ; (5) New South Wales, 19.33 ; (6) Tasmania, 18.53 ; (9) South Australia, 17.99 ; (8) New Zealand, 16.65.

Middle-aged Men and Women, 30-50 years.—(1) Queensland, 23.60 ; (2) Western Australia, 22.63 ; (3) New South Wales, 22.61 ; (4) South Australia, 21.82 ; (5)Australasia, 21.75 ; (6) New Zealand, 21.67 ; (7) Victoria, 20.47 ; (8) Tasmania, 20.21.

Elderly Men and Women, 50-65years—(1) Victoria, 9.93 ; (2) New Zealand, 8.94 ; (3) Western Australia, 8.82 ; (4) Australasia, 8.50 ; (5)Tasmania, 7.91 ; (6) South Australia, 7.82 ; (7) New South Wales, 7.54 ; (8) Queensland, 7.05.

Aged Men and Women, 65years and over.—(1) Tasmania, 4.60 ; (2) Victoria, 3.45 ; (3) South Australia, 3.32 ; (4) Western Australia, 2.95 ; (5)Australasia, 2.81 ; (6) New South Wales, 2.52 ; (7) New Zealand, 2.30 ; (8) Queensland, 1.52.

Males, 20-40years, Soldier's age.—(1) Western Australia, 40.36 ; (2) Queensland, 38.52 ; (3) Victoria, 34.52 ; (4) New South Wales, 34.47 ; (5)Australasia, 33.77 ; (6) South Australia, 32.35 ; (7) Tasmania, 32.05 ; (8) New Zealand, 28.46.

Females, 15-45years, Fertile age.—(1) Victoria, 48.55 ; (2) Queensland, 46.66 ; (3) Australasia, 46.58 ; (4) New South Wales, 46.42 ; (5) Western Australia, 46.34 ; (6) South Australia, 44.78 ; (7) New Zealand, 44.72 ; (8) Tasmania, 44.72.

SECTION V. CONJUGAL CONDITION OF THE PEOPLE, 1891.

Of the total population, 146,667, particulars relating to the age and conjugal condition of 146,025 persons, or 99.57 per cent. of the whole, have been ascertained. The proportional analysis of these may therefore be safely taken as applying to the total population.

The following summary gives under particular age-groups a sufficiently comprehensive view of the number and percentage proportion living under the several conjugal conditions :—

From the above analyses it is apparent that for every 100 persons of all ages there are—

If we restrict the relative proportions to all married and marriageable persons, however—say, males 20 years and upwards and females 15 years and upwards—we obtain the following proportions in relation to all married and marriageable persons; thus :—

Marriageable Men and Women.

To ascertain the full extent of marriageable persons it is necessary to add the widowed and divorced persons to bachelors and spinsters, as in the following abstract :—

Although there were, in 1891, 104.18 marriageable women to every 100 marriageable men, it appears there were only 95.44 spinsters to every 100 bachelors. In 1881 there was a larger proportion of marriageable women in relation to marriageable men, being 107.10 in 1881 to 104.18 per cent. in 1891.

Distribution and Proportion of marriageable Women to marriageable Men, 1891.

As usual the larger number of marriageable women relative to marriageable men are found in the cities, and the smallest in country districts, especially so in the partially settled mining districts ; thus :—

The smallest proportion of marriageable women to men was in the mining district of Cumberland, viz. 22.05 per cent. ; and the largest in Launceston South, where it was as high as 210.35 per cent.

Proportion Married and Single Persons of all Ages at the last Seven Census Periods.

The proportion of married and single at various age-periods can only be ascertained for the two last Census periods. The proportion for all ages at the last seven Census periods are shown in the following abstract :—

SECTION VI. BIRTH-PLACES OF THE PEOPLE.

Of the total population at the last Census (146,667), particulars regarding the place of birth of 146,043 persons were ascertained. Of these, 143,104, or 97.99 per cent., were born in British possessions, and 2939, or 2.01 per cent., in Foreign countries. Those born in Australasia numbered 115,229, or 78.90 per cent., of which 107,901, or 73.88 per cent. were born in Tasmania. Those born in the United Kingdom numbered 26,975, or 18.47 per cent., of which 17,443, or 11.95 per cent., were born in England and Wales ; 3845, or 2.63 per cent. in Scotland ; and 5687, or 3.89 per cent., in Ireland. The extent to which persons born in different countries have increased or decreased during the three last Census periods is more fully shown in the following table :—

Birth-places of Adults.

If the number of adults born in different countries are compared, the per-centage for Tasmania will be lessened considerably, while the per-centage proportion for the United Kingdom and other countries will be greatly increased. This is shown in the following abstract comparing the proportions of adults and all ages :—

Mean, Increase since 1881.

Of the means increase, 30,962, to the population since the Census of 1881, 31,251, or 101.77 per cent., were born in the Australasian Colonies. There was a decrease of 1505, or 4.90 per cent., in respect of persons born in Ireland. In all other countries there was an increase of 3.23 per cent. A more detailed statement of the increase or decrease since 1881 in respect of the persons born in the different countries are given, as follows :—

Tasmanians in each Australasian Colony.

There were 23,936 Tasmanians living at the last Census in the other six Australasian Colonies. These, with the 107,901 enumerated living in Tasmania, make a total of 131,837 persons of Tasmanian birth living in the Australasian Colonies. The other Colonies have been the gainers by 16,627 persons in migratory exchange, as there are only 7328 persons living in Tasmania who were born in other Australasian Colonies. The following abstract gives the details of the distribution of the 131,837 persons of Tasmanian birth living at the last Census in the Australasian Colonies :—

SECTION VII. EDUCATION OF THE PEOPLE.

It is only in a very narrow and restricted sense that statistical information gained by the ordinary Census methods can even be accepted as an index of the "Education of the People." At the very most it only attempts to gauge the proportion of those who have acquired the power to receive or impart ideas to others by means of at least one known system of written or printed symbols. Of the extent, nature, or value of the ideas themselves, or of the power to receive or impart them in perfection in one or many symbolic forms, as tested by the varied educational standards of science and philosophy, the educational statistics of the ordinary Census tables do not afford the slightest clue.

This is best appreciated by asking how far the knowledge of the mere ability of any one to read and write is sufficient evidence of the possession of such a standard of education as would satisfy the Industrial Chief, the Statesman, the Schoolmaster, or the Philosopher. Of the usual subdivision of Education into Physical, Intellectual, Moral, Religious, and Technical, the power to read and write conveys little or no information.

Notwithstanding these objections, there is much to be gained by the knowledge obtained from the statistics which disclose the proportion of persons who can read and write at each age-period from 5 years and upwards ; and valuable information is also afforded in respect of the means provided, and the extent to which these are availed of at school-age, in every part of the country. Thus although, directly, the comparative power to read and write at each age and in each district conveys little information of the nature and extent of the education imparted, it is eloquent as an index in revealing comparatively how far the ordinary but valuable culture of the Common School system extends or is lacking ; and also it affords a very safe measure of the success attained in other branches of education which are the usual concomitants of tuition in reading and writing.

As the power to read and write, therefore, is regarded mainly as evidence of the possession of a certain amount of intelligence obtained from the training in all other matters usually comprehended within the Common School curriculum, it has been deemed advisable in this report to include all foreigners who can read and write under this category, irrespective of the tongue or symbol used, as there is no more justification, on educational grounds, for excluding foreigners who can read and write only in their own tongue, than there would be in excluding Englishmen who can only so express themselves in their mother tongue. In the detail tables, however, the numbers of those who can only read, or read and write in a foreign tongue, are given separately, so that comparisons may be made with other places where a different method of classification obtains.

In making comparisons it is obvious that all infants and children not yet arrived at the age for training in reading and writing should be excluded. Indeed, it is the results of the training at the end of the normal tutelary stage which should be sought for in comparison with other periods or countries, rather than the misleading averages, which include not only the young under and during the age of the process of training, but also aged persons, whose low averages tend to conceal the success or otherwise of the existing machinery provided for the education of the young of the community.

The classification adopted at the last Census by the Statisticians of Australasia, for obvious reasons, excludes all ages under 3 among those stated who could "read" and all ages under 5 among those stated who could "read and write." In phrase "for all ages," used hereafter in tables or otherwise, it must be understood that the young under the categories referred to are wholly excluded.

The following table shows the percentage proportions of all who could read and write, read only, and could not read, at the principal groups, as ascertained at the last Census in Tasmania :—

From the above very full analysis it is shown that the percentage proportions of all over the School-age—14 years and over—under various categories, were as follows :—

The proportion at the higher stage—read and write—at each quinquennial period is highest at age-period 15-20, among which Females (97.00) per cent.) show the best results, thus—

As the percentages of those who can read and write at successive quinquennial age-periods gradually diminish until it is as low as 55.68 per cent. in the oldest (85 years and over), they afford the best indication that can be given of the decided but gradual improvement in the means of education that has taken place at each quinquennial since the beginning of the present century ; and that even the present means of education is capable of producing still higher

results, as years eliminate the poorer results and reduce the proportions of past efforts, is shown by the figures which record more perfectly the results of the educational system now in operation—that is at the age-period 15-20 years—among which those who cannot read and write do not exceed 5 per cent. of those living at that age.

Improvement in Education since 1881.

The tendency towards improvement is also clearly manifested in comparisons at different age-periods, as shown by the Census returns of 1881 and 1891 ; thus—

Education of Children.

The education of children between 5 and 15 years of age upon the whole shows considerable improvement, as compared with the last Census (1881) ; thus—