(1) liability to insanity were independent of age once the age of twenty is reached, and as if (2) the age-distribution for ages above twenty were the same for both the native-born and immigrant populations. Now, this is very far from being the case in either particular. For instance, at the ages 20-25 there are only 13.32 lunatics to 10,000 of population; at 65-70 there are 133.34, an increase in the ratio of 1 to 10; and the increase is steady from the one age-period to the other. Again, in the former of these age-periods the native-born population is more than five times the not-native-born; in the latter it is less than one-seventieth; and again there is a steady change from the one position to the other in between the two age-periods. Fig. 1 represents by graphs the age-distribution of the whole population and also that of the native-born. It represents also the lunacy-rate—that is, the number of lunatics to 10,000 of population at each age-period. The diagram shows clearly how, as the lunacy-rate increases, the ratio of the native-born to the whole population of the same age-period becomes less and less, until by the time the lunacy-rate has reached its maximum this ratio is very small. Consequently, in the two populations of over twenty, the native-born are to a far greater degree than the others concentrated in the earlier age-periods, in which the liability to lunacy is much smaller than in the later ones, and consequently the number of native-born lunatics is to be expected to be very much less than it would be if with the same populations these differences did not exist. We have further to allow for the population not New-Zealand-born containing a large majority of males, with their greater liability to insanity.

In Table I, I have made an attempt to allow for these differences. The table gives the whole population and then the native-born population in quinquennial age-periods. The fourth and fifth columns give the latter divided into males and females. The next two columns give the number of lunatics per 10,000 of population for each age-period. On the hypothesis of equal liability to lunacy at the same ages in both classes of population it is now a simple matter to calculate the proper number of native-born lunatics of each sex in each period. The results are entered in the last two columns.

This method, it should be noted, still exaggerates the estimate of the number there ought to be of native-born lunatics if they had the same liability to insanity as the remainder of the population. For the condition of things we are trying to allow for still holds for each of our age-periods. The proportion that is native-born is greater in the earlier than in the later years of each age-period, and up to about seventy years the liability to lunacy is continually on the increase as the age increases, while after this age the contribution to the number of lunatics is small by reason of the smallness of population. It is impossible to allow for this feature with perfect accuracy. If we took yearly age-periods the error would be smaller, but the problem is hardly worth this, and it is sufficient for our purposes that it should be clearly understood that the estimate in the table is distinctly an overestimate. The figures in the table are all taken from or based on the census returns for 29th April, 1906. The lunacy-rates quoted here and elsewhere are those for New Zealand as determined by this census, but their general character is in no way peculiar. Their main feature is general—namely, an increase in the rate up to sixty-five or seventy years of age, after which there is a not very great reduction, due to the greater mortality of he insane causing a reduction in their relative numbers.

the average of the country of his origin.” This is said first of all to be suggested by the comparison of the immigrant with the New-Zealander, in which it is pointed out that the insane of the immigrant population above the age of twenty is as great as 1 in 118-9. After the inquiry in the preceding section of this paper, this suggestion would appear to cease to have any force. It is then stated to be verified “by contrasting our figures with the English figures (only approximate) of 1 insane person below twenty to 2,069 of the population of the same age, 1 in 173 of the population above that age, and 1 in 283 of all ages.” Now, in arguing from these bare figures serious mistakes have been made similar to those made in contrasting the New-Zealand-born with the immigrant. The remarkable age-distribution of those of our population not born in this country, combined with the increasing liability to insanity as age increases, is not taken any account of above the age of twenty.

To illustrate the extent of this peculiarity that has to be allowed for, let us compare the age-periods 20–25 and 45–50. The number of lunatics per 10,000 in the later period is more than six times as great as it is in the earlier. Our total population in the later period is about two-fifths of that in the earlier, whereas our immigrant population of the later period is almost double that of the earlier. Further, whereas in the period 20–25 our immigrant population only constitutes one-sixth of the whole population of that age-period, in the later age-periods it practically constitutes the whole population of the periods. We have further to take into account the very considerable excess of males in our immigrant population, amounting to nearly 50 per cent., and their greater rate of insanity, and also the excess of females in older countries. In fig. 2 we have the graphs representing the population of New Zealand not New-Zealand-born. The excess of males stands out clearly, and a comparison of these graphs with those in fig. 1 for the whole population reveals clearly to the eye the exceptional character

of the age-distribution of those not born in New Zealand. The graphs representing the rates of lunacy for the two sexes are also repeated in fig. 2, and two other graphs are given representing how our population above twenty years of age and not New-Zealand-born would be distributed in age and sex if such distribution were the same as that of the population of England and Wales as given in the census of 1901; or, in other words, representing the age and sex distribution of a portion of the population of England and Wales over twenty equal in number to that of our population over twenty not New-Zealand-born. A comparison of these graphs with the others in the same figure again reveals clearly the exceptional character of the age and sex distribution in question. It is obvious, then, that, in view of the very different rates of insanity prevalent in the various age-periods, the figures as presented by Dr. Hay cannot be accepted as any indication of the comparative insanity of the immigrant and his countrymen at home.

In Table II I have made an endeavour to roughly measure the correction that must be applied to allow for the features I have pointed out. The first column gives the age-periods beginning at 20-25, the next two give the population of New Zealand according to age-period and sex, and the next two again our immigrant population classified in the same way. The sixth and seventh columns give our total immigrant population above the age of twenty redistributed in respect to both age and sex proportionally to the population of England and Wales according to the census of 1901. The next two columns again give the New Zealand rates of lunacy, which have been already shown to apply generally to the New-Zealand-born and the immigrants as a whole. The last two columns give the number of immigrant lunatics there would be if our immigrant population of twenty years and over were distributed in age and sex as the population of England and Wales is distributed.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

The result is to give us 1,589 lunatics, to which ought to be added the share of idiots to get the total insane for comparison with English figures. But, even ignoring these, this is at the rate of 1 in 164.5—a very considerable change from the actual 1 in 118.9, and approaching pretty closely to the 1 in 173 of England and Wales which Dr. Hay takes for comparison, especially when we remember that the dealing with quinquennial age-periods does not make by any means a complete correction.

The design of Table II is not perfect, but perfection cannot be attained with the statistics available. The immigrant population contains large numbers born in Scotland, Ireland, and Australia. Those born elsewhere are relatively so few as to make no material difference in the argument. The Scotch differ from the English only slightly in respect to liability to insanity, and the population of Scotland is distributed with respect to age very similarly to that of England and Wales. Those of English, Welsh, and Scotch birth form the great majority of our immigrant population. Those of Irish and Australian birth are not very unequal in numbers—their contributions to the insane when combined are about equal proportionally to that of the other nationalities—and the age-distributions of the populations of Ireland and Australia vary in opposite directions from that of Great Britain, Ireland having a greater proportion of older and Australia of younger people. So the table does not altogether fail in taking the immigrant population of New Zealand and finding what the position would be if their age-distributions were similar to those of the native countries of the immigrants. The inference is that there is no appreciable evidence that in the total the immigrant is any more liable to insanity than the people of his native land; and, even if there were in the mere statistics, the question would arise to what extent the apparent differences were really due to differences of classification and of circumstances in the different countries leading to more or less perfect returns, and differences of asylum accommodation and treatment of the insane leading to greater or less longevity.

and over, redistributed in respect to both age and sex proportionally to the census of 1901 for England and Wales.

The question now arises, Will these differences fully account for the deficiency of female lunatics as compared, say, with England and Wales? Some attempt has been made to answer the question in Table IV. In this table is calculated the number of male and female lunatics over twenty years of age there should have been returned at the various age-periods in the census of 1906 if our population over twenty had been distributed in age and sex as was that of England and Wales at the census of 1901.