food, clothing, and heat, with more immediate injurious results than even a dark room or absence of bath.

7. Income groups $10-$15, $16-$19 are not so fair for purposes of comparison as would be $10-$14 and $15-$19.

The only use made of these income groups is to point out that physical defects are found among all of them, and that defective home environment is common to all. No attempt is made to compare group with group except in a summary table (XXXVII.), which we printed merely for the purpose of laying all the facts before students who might wish to compare group with group. Table XXV. gives the units upon which Table XXXVII. is based.

8. Income units vary, being one dollar for incomes under $10; two dollars for incomes over $10 and under $40; five dollars for incomes over $40 and under $50; twenty-five dollars for incomes over $50 and under $100.

These divisions were made because they best served the purpose of the Committee's study. No practical and no statistical purpose could be served by reporting incomes by dollar units where the weekly return is over forty dollars. There was no way of knowing in advance that important differences would not show between the $5 and the $6 group, or between the $10 and $11 groups. The $1 unit was used for all cards up to $25. In tabulating, it was seen that no useful purpose would be served by separating $11 from $10 and $13 from $12. The only reason that Table XXV. giving the two groups was published was to make available material for later statistical inquiry.

9. The term "per cent. distribution" in Table XXXVII. is not clear. This we regret. While our 20 significant facts and 15 practical conclusions were intended to show that physical defects were common to all income groups, and that, therefore, remedies must reach causal conditions among all income groups, we realize that, given equal care, children of lowest income groups are more likely to have physical defects and home handicaps than children of adequate income groups. 100 seemed the best common denominator available. We therefore set up a consolidated comparative table that would show whether or not the proportion of defects and unfavorable conditions was higher or lower in a given income group than its proportion to total families. By failing to separate from the general table the two summary columns $10 to $19 and $20 to $29, and by failing to set up a final 100 per cent. column, by which to check additions of the six income groups, we failed to make the most out of this table. Nevertheless, the table brings out certain significant facts. For instance, it shows that, while among 1,400 families visited only 8.4 per cent. had weekly incomes less than $10, this same small group con

tained 13.8 per cent. of the malnutrition, 9.6 per cent. of the defective breathing, 10.3 per cent. of the families losing children, and 8.6 per cent of those paying over 25 per cent. of their income for rent. It is also inter. esting and suggestive to note that, while families having an income of over $30 a week are but 15.6 per cent. of the total families, they include 27.6 per cent. of those who pay over 25 per cent. of their income for rent.

Permit us to express our appreciation of the service rendered by the American Statistical Association in presenting at a time of growing public interest in school hygiene this study of the home conditions of school children suffering from physical defects.

COMMITTEE ON PHYSICAL WELFARE OF SCHOOL CHILDren.

NOTES AND REVIEWS.

NOTE ON PROPOSED RULES OF STATISTICAL

PRACTICE.

The last issue of the Publications contained (page 523) some proposed resolutions relating to the practical collection and compilation of vital statistics, the purpose and scope of which may not be entirely clear without some knowledge of the character and organization of the body to which they were presented; namely, the newly organized Section on Vital Statistics of the American Public Health Association.

As indicated in the Publications, the proposed rules formed an Appendix (Appendix C) of a paper, from which is extracted the following explanatory statement:

Third, and last, of our most pressing general requirements would seem to be the establishment of uniform Rules of Statistical Practice, which should be carefully formulated, supported by evidence of their necessity, carefully considered by the special committees of reference and by the registration offices which are to carry them out, and when finally adopted under the auspices of this Section and of the American Public Health Association, then they should be strictly complied with by all of the registration offices of the United States. Otherwise, practical uniformity and comparability of vital statistics will be quite out of the question. Every safeguard should be employed against hasty or ill-considered judgment, but, once settled, a Rule of Statistical Practice should govern absolutely until repealed. As examples of some questions on which we should legislate, some proposed resolutions are submitted (Appendix C) on the following subjects:

1. Statement of cause of death.

2. Statement of occupation.

3. Statistical definition of deaths.

4. Statistical definition of stillbirths.

5. Statistical definition of births.

6. Essential requirements for registration of deaths (reindorsement). 7. Essential requirements for registration of births.

8. Method of testing accuracy of registration of deaths. 9. Method of testing accuracy of registration of births. 10. Constitution of standard tables of vital statistics.

11. Adoption of uniform age-periods in mortality statistics. These proposed rules, together with those that may be suggested by other members of this Section, will probably afford sufficient material for committee work during the ensuing year. By keeping an accurate record of the decisions upon all propositions, whether pro or con, we shall

soon have a valuable code of statistical procedure to guide our action, and will know as well what has not been approved and so avoid statistical byways leading to confusion.

The resolutions are now under consideration by the various committees of the Section on Vital Statistics, and their recommendations for action in regard to some or all of the propositions may be submitted at the next annual meeting of the Association at Winnipeg, Manitoba, August, 1908. Suggestions or recommendation relating to the subjects covered, or to other matters of statistical practice, will be warmly welcomed from members of the American Statistical Association and from all others interested. They may be sent to Dr. Wilmer R. Batt, State Registrar, Harrisburg, Pa., who is Secretary of the Section on Vital Statistics, and who will refer them to the appropriate committees.

CRESSY L. WILBUR.

MUNICIPAL STATISTICS.

Statistical Summary of the City of Buenos Ayres. By Albert B. Martinez, Director of Municipal Statistics, 1906.

This report is the sixteenth of its kind, the bureau of municipal statistics of Buenos Ayres having been established in 1881 by Dr. Emilio R. Coni. The present volume contains an introduction of 55 pages and the main part of the work 352 pages. The introductory part is of value to all students of vital statistics, as it contains up-to-date international comparisons of birth, marriage, and death rates. These latter distinguish the mortality of children under one year of age. There is also an international comparison of eleven infectious-contagious and other causes of death on the basis of ratios of the specified causes to all causes. Some of the facts presented in the statistical summaries are of quite exceptional interest, and are seldom, if ever, found in the municipal reports of the cities of the United States. Under climatology, the barometric pressures; temperature and relative humidity; and the psychometer readings, dry and moist, are given for three daily readings,-7 A.M., 2 P.M., and 9 P.M.,-the averages being given for every ten days or as nearly as this is possible, to give three averages for each month. The rainfall and average velocity of winds are also given for the thirty-six divisions of the year. The heliometric observations are presented for each month, and include the hours that the sun was above the horizon, the hours that it was visible in the morning and in the afternoon, and the number of cloudy hours. The atmosphere is chemically analyzed for ozone, during the day and during the night, for carbonic acid, and for organic ammonia. Finally,

*The exact date of the meeting is not yet determined.

the air is bacteriologically analyzed for bacteria and what is called "mustiness." The results of the chemical analyses are averaged by months, but the bacteriological results are expressed in periods of from three to eight days' duration of bacteria growths from samples of air taken near the first and middle of each month.

Of other more or less unique statistics, mention may be made of the statistics of race tracks, or, as they are locally called, "hippodromes." These tables show, among other facts, the amount of money wagered on the horses. They show that race-track gambling has rapidly increased in Buenos Ayres, and during 1906 over $49,000,000 was wagered at the "hippodromes," or an average of nearly $50 per capita for the city's total population. Theatre statistics, financial statistics of the National Charitable Lottery, strike statistics, and returns of the Sick Fund Societies are included in other tables of more than local interest.

Under the heading "demography" there are some valuable statistics of births, marriages, and deaths. Births are returned as legitimate, illegitimate, and still-born, though the latter are not considered true births. Twin and triplet births are separately tabulated, as are also the illegitimate births legitimatized by declaration of one or both parents. The legitimate children are classified by nationality of mothers and order of birth. This table shows, for example, that in 1906 two Italian mothers gave birth to the twentieth child. The order of birth is also tabulated by ages of the mothers. In 1906 there were 1,996 children born of mothers under twenty years of age and five children born of mothers over fifty years of age. Of the former number, 1,467 were first births, 424 second, 74 third, 4 fourth, 4 fifth, and 2 sixth; while of the latter one was a second child, one a sixth, one a seventh, one an eighth, and one a twelfth.

The marriage statistics are given in less detail than the births, and are principally interesting because of the ages. The most popular age for marriage is under twenty for girls. About one-third of the young men marry at ages 21-25 and another third at ages 26-30.

The mortality statistics occupy thirty-eight pages of the volume, and the tabulations are in many respects admirable. The crude death-rate of Buenos Ayres has declined from 27.2 per 1,000 in 1888 to 16.5 in 1906. The rate was highest (30.0) in 1890 and lowest (15.5) in 1905. Tuberculosis seems to have been somewhat more fatal to the native Argentines than to persons from outside of the Republic. During 1902-06, among those of Argentine birth of ages fifteen and over, 28.5 per cent. of the deaths from all causes were from tuberculosis as against only 14.3 per cent. for persons from outside of the Republic.

The mortality of children is an important subject admirably treated in these statistics of Buenos Ayres. Although the divisional periods of life at ages under five are not given in the detail available in the statistics of