headed Area,' upon which we cannot help remarking, that though the Commissioners of the Census have availed themselves without alteration of the slightly erroneous admeasurements of the counties which had already answered the purpose of Mr Rickman, with whose returns these were to be compared, and though they have not thought it right to order new surveys at an expense not contemplated by the Act; yet we feel that the time is come, when this omission in the very A B C of our Statistics should be supplied. We would gladly also suggest the removal of certain inconvenient, but ancient anomalies of local division, by which portions of townships, parishes, and even counties, appear as if dropped into the midst of their neighbours' share in the very process of distribution. We do not desire to remodel to too great an extent old boundaries; we have no wish to see the princely proportions of Yorkshire cut up into the matter-of-fact circles of a French map; but we cannot help feeling that, in these days of more sober sense, a mere romantic attachment to ancient divisions will not, and ought not, to prevent the removal of palpable anomalies or continually recurring inconveniences. But this subject, though important and deserving of immediate attention, would draw us on too far at present.

Whatever difference of opinion, however, there may be as to remodelling old geographical divisions, there can be none, we presume, as to the necessity for making such divisions (while they still exist) the foundation of all statistical returns; still less as to the impropriety of permitting two returns so intimately connected together as those of the Census Commissioners and of the Registrar-General, to refer to districts which, though bearing the same name, are very far from conterminous. We shall, we are apt to think, be hardly credited when we state this to be the case with the two publications before us on the population, and on the births and deaths in the English counties. Let us make our meaning clearer by an example. Any enquirer would suppose, that by taking the total number of births or deaths in one year for a county, such as Worcester, or a town, such as Kidderminster, as given by the returns of the Registrar-General, and dividing them among the total population as returned by the Census Commissioners, the result would give correctly the proportionate number of births or deaths for that county, and town, respectively. This is a proposition so obvious, that some surprise will be felt when we add, that such an arithmetical comparison would only lead to error, and has in fact already done so in documents before Parliament, founded on a computation such as we have supposed.

The truth is, that at the time the General Registry of births,

marriages, and deaths, was introduced, the limits of poor-law unions were unfortunately directed to be those also of the registry districts-districts no doubt convenient in extent, and already arranged with reference to the various important considerations of amount of population, and facility for being comprehended under one authority, but alas! not conterminous with the ancient geographical divisions of England, as regards counties and large towns. To the Registrar-General's returns of births, deaths, and marriages, in the several counties, is appended the following note: "When the districts run into two or more counties, they have been classed in those counties in which the greater part of the popu، lation was situated; hence these groups of districts rarely, if ever, 'correspond with the strict boundaries of the respective counties.' Here is fair warning given to those who happen to read the note; but still, as these ، Registry Counties' (to give them a distinguishing title) bear the same name as the recognized county divisions of England; and as returns of the births or deaths in them, are headed births in counties' and 'deaths in counties'-the population given by the Census returns to those counties, has been, in many statistical works, taken as that among which the registered births and deaths are to be divided in order to get at comparative results.

،

That such a process must lead to the greatest errors, will be immediately perceived. In the very interesting Report on the sanitary state of the labouring classes, prepared under the superintendence of the Poor-Law Commissioners, and ordered by Parliament to be printed, a very curious table occurs.-(P. 183.) It is intended to exhibit, at one view, the proportions of births and deaths to the population for every county in England, in order to show that an excessive number of births and deaths usually go together, and thus mutually compensate each other; and secondly, to exhibit in certain counties considerable increase of population in spite of very great mortality. Now in this table, if we turn to the county of Worcester, we shall find the deaths set down at 1 in 33-a mortality for a whole county not only infinitely above the average mortality of England; but nearly as great as that of the most crowded manufacturing towns in their least healthy quarters. The births, on the other hand, are given as 1 in 20-a rate of fecundity the more startling, as it would appear unparalleled in any other spot in the kingdom. Various are the causes which have been suggested in Statistical publications, to account for this excessive mortality-reasons at least ingenious, if not well founded. Without, however, using the coarse but emphatic expression adopted by Paley, with reference to some other equally puzzling anomaly—' It's a lie, and that's the

solution-we may content ourselves by stating, that the fact is not so, and that the actual mortality is very little higher than 1 in 50; the number of births, 1 in 29.

This mistake (and it extends to most of the other counties) has arisen from the registration districts, which are concentrated under the title of Worcestershire, extending into the counties of Stafford, Warwick, Salop, Gloucester, and Hereford; and therefore embracing a population larger than that which has been taken as the basis of calculation, by 102,624; that is to say, an addition of nearly fifty per cent on the original population of 233,484. When it is remembered that this increases the apparent proportion of births and deaths to the population by one half, the excessive mortality and fecundity of this unfortunate county is at once explained. We need hardly add, that all the births and deaths which thus unduly swell the totals for Worcester, must also diminish those of the surrounding counties, which, in return, borrow from their neighbours; till the whole value of the Census and Registry abstracts, for mutual comparison, is destroyed. We find, in fact, upon examining into the details, that while Worcester takes the greater part of its excess of 102,000 from the county of Stafford, that county is in some degree recompensed (though still left deficient to the extent of 67,000) by taking nearly 20,000 from Derbyshire, which latter revenges itself on 12,000 good men of Nottingham and Leicester. The counties in which, next to Worcestershire, the greatest changes of a similar kind have taken place, are, Derbyshire, which is thus diminished one-third-Berkshire, increased by a fifth; then come Nottingham, Monmouth, and Lancaster, of which we will spare our readers the details. We will only add, that of all the counties of England and Wales, with the solitary exception of Cumberland, there is not one which has not undergone more or less transmutation. If the evil were restricted in its consequences to the mistakes into which calculators might fall-from ignorance of the real limits and population of the registration counties-the inconvenience, though great, would admit of the remedy of furnishing the exact population of each registration county; but there is this additional drawback, which is inseparable from the system, that the whole character of the county is in many instances changed. Counties which before were wholly agricultural or commercial, receive an admixture of population, which either reverses their character, or gives them that mixed complexion which weakens every distinctive feature of the statistical facts applicable to them, and destroys the effect of comparison with former ascertained results. If we find that the per centage of mortality for an agricultural county, with some admixture of manufactures, (such as

Worcester was in former times,) is now increased-how can we form any just conclusions from such a result, without knowing the exact character of the 100,000 persons who have been added to it from the neighbouring counties? We might multiply instances as to counties; but if the inconvenience from change of character is so great with these, still more so is it with the returns of births and deaths in large towns-in some of which, the district of the same name comprehends a population distinct from, and nearly double that of the town after which it is called; and yet the return of births and deaths for such district, is the only medium afforded us by which to measure the rate of mortality or fecundity in the town which is the subject of enquiry.

Thus, if from finding that Kidderminster forms a remarkable exception to the general rate of increase in large towns, (having actually decreased four per cent in the last ten years,) we should wish to see whether there is any thing in the proportionate relation of births to deaths that can account for this-we examine the columns of the Registrar-General, and find a separate return for Kidderminster; but in order to avoid falling into the mistake which we have already pointed out, we look to see the amount of population to which it refers, and find, that whereas the town of Kidderminster contains a population of 14,399, the Kidderminster district contains 29,408. How far any peculiarities in the rate of mortality of the town, about which we are enquiring, may be modified by an additional population, equal in amount to its own original number, we have no means of judging, and therefore give up that enquiry in despair. In like manner, for the borough of Leeds, containing 152,054 persons, we have no separate return, but only for the district of Leeds, containing 168,667. For Bradford, in Yorkshire, of which we might wish to test the nature and causes of the extraordinary increase of forty-eight per cent in ten years, we are left to infer the births and deaths among the present population of 34,560, from those of the registration Bradford, which apply to 132,164. We should have liked to have seen how much of the marvellous stride in numbers which West Bromwich has made since the last Census, viz. seventy per cent, arises from immigration, how much is a natural increase from excess of births over deaths; but to say how many of the births and deaths returned for the West Bromwich which embraces 52,596 inhabitants, may be allotted to the West Bromwich which contains but 26,121, would require some peculiar faculty, which we pretend not either to name or possess.

In the Fifth Report of the Registrar-General, Mr Farr has made allusion to this effect of the construction of the registration districts, in as clear terms as could reasonably be expected from a sensible man, in describing a system under which he is acting.

6

After comparing the mortality exhibited by equal populations in country and in town districts, he adds with respect to the latter, the registration districts are generally single parishes, or pa'rishes united for the relief of the poor; and were formed by the poor-law commissioners with this object exclusively in view; hence the twenty-five town districts comprise entire parishes, which have all the character of country districts.' He adds, the effect of the agents is understated, by which the disease and mortality of towns is caused. It is as if the specific gravity of two masses of metal were compared, the one containing eight, the other two parts in ten of gold; the ratio of the specific gra'vities of the two masses would be less than the ratio of the specific gravities of the pure metals, but a comparison would prove incontestably, that gold was by far the heaviest.'* True; but no one who wished to compare the specific gravities, would of his own free choice employ alloyed metals for the purpose; still less if (as in the case before us) there were no certain mode of testing the amount of alloy. Would it not be much more satisfactory, if, in comparing the mortality of town and country districts, we could arrive at actual and precise, instead of merely approximate, results?

If we have pursued this subject at greater length than may be agreeable to some of our more impatient readers, we must plead in excuse our strong desire to see some remedy applied without loss of time; and also our repeated disappointments at the failure of all our efforts to draw conclusions from a comparison between these two depositaries of statistical information ;-disappointments which have been solely caused by the difference of the form in which they are given, and the districts to which they apply. They are at present like two well-educated individuals furnished with varied information upon an interesting topic, but who have no chance of ever meeting; or, to use a mechanical simile, like a railway train and its engine which have got upon different lines of the same railway, so that there is excellent moving power side by side with ample accommodation for passengers, and yet the traveller is unable to proceed on his journey, Our object is to smooth away the difficulties that prevent the meeting of two such well-informed gentlemen, and get, if possible, the engine and its train upon the same line.

Nor have we thus dwelt upon the evil, without being prepared to suggest a remedy of immediate application. It is true the Census returns might on the next occasion be collected, and all its various heads of information arranged for the registration dis

*Fifth Report of the Registrar-General, p. 201.