bought and sold by or subject to arrangement of local board, and become private property. Districts supply indigents.

Illinois.-District board fixes list for four years. Indigents supplied free. Kentucky.-County board of examiners fixes list for five years with publishers' guaranty. The county judge furnishes indigents.

Louisiana.-State board fixes list for four years with limited local discretion. Mississippi.-The county school board adopts a series of books for five years on publishers' guaranty. Penalty pupils without the prescribed books in any branch are not to receive instruction in that branch.

Missouri.-A State school book commission fixed a list, with publishers' guaranty, for five years from September 1, 1897, to be handled through dealers. Indigents are supplied from local contingent funds.

Montana.-Local boards prescribe studies additional to those in State law, and furnish indigents.

Nevada.-State board fixes list for four years. District furnishes indigents. New Mexico.-Territorial board of education is authorized to fix a list for four years, and to contract with publishers and sell to counties. Districts furnish indigents.

North Carolina.-County board fixes list for three years, with publishers' guaranty.

Ohio.-A State commission fixes a list, on publishers' guaranty, from which local boards fix lists for five years (with exception). Boards may buy and sell to pupils or arrange with dealers to supply them. Indigents are furnished.

Oklahoma.-Territorial superintendent fixes a list for five years on publishers'

guaranty.

Oregon.-State board fixes a list for six years on publishers' guaranty. Penalty, forfeiture of apportionment.

South Carolina.-State board fixes a list for five years on publishers' guaranty, and may require publishers to have depositories in each county.

Tennessee.-County superintendent suggests suitable books.

Texas.-A law resembling that of Missouri is to take effect September 1, 1898. Penalty upon any teacher or trustee, $10 to $50 for each offense. Every day of violation of law to be considered a separate offense.

Utah.-A convention of superintendents fixes a list for five years on publishers' guaranty. Penalty on teacher, forfeiture of eligibility.

Virginia.-Two books of John Esten Cooke (Virginia, a History of her People; and Stories of the Old Dominion) are prescribed by law. State board fixes a list. Washington.-State board fixes a list for five years on publishers' guaranty. Penalty on district, loss of one-fourth of the apportionment. Local boards furnish indigents.

West Virginia.-A contract list for five years is part of the law of 1896, with exceptions. County schoolbook boards are established by act of 1897. Publishers keep books with local depositories on account of district building fund. Penalty on every officer or teacher, $3 to $10 for each offense,

Wyoming.-Same as Utah, except penalty.

The States following, regularly or through stated action, authorize provision for free use of books by pupils:

Colorado.-District boards fix list for four years, with exceptions. Indigents are furnished, and, on popular vote, all pupils, free.

Connecticut.-State board may fix list for five years. Town boards may take additional action and, on popular vote, furnish free text-books.

Delaware.-State board fixes list; district board furnishes free text-books. Idaho.-Books adopted by a State board of text-book commissioners for all common, graded, and high schools, are furnished free by the district under contracts with publishers for six years, ending September 1, 1899.

Iowa.-Local boards may buy and sell to pupils at cost. County uniformity can be fixed for five years. Text-books are furnished free to indigents, and, on popular vote, to all by the district.

Kansas.-A school text-book commission (1897) has selected text-books in common school studies for five years, and contracted with publishers to furnish them to pupils through agencies at every county seat. On popular vote, with a twothirds majority, school boards may purchase books and furnish their use free to pupils. Penalty for using other text-books, except for reference, $25 to $100, with or without imprisonment.

Maine, New Hampshire, Massachusetts, Rhode Island (towns); New Jersey, Pennsylvania (local boards); and Maryland (counties), furnish free text-books. Michigan.-District boards furnish books to indigents, and, on popular vote, to all pupils free.

Minnesota.-Local boards may fix a list for three to five years with publishers' guaranty, and may purchase and provide for loan free or for sale at cost to pupils. Nebraska.-Local boards furnish books free, and may fix list, with publishers' guaranty, not beyond five years. A local dealer may be designated to handle the books on agreed terms.

*

*

New York.-Every union free-school board is "to prescribe the text-books and to furnish the same out of any money provided for the purpose." Commonschool districts, by popular vote, may furnish indigent pupils.

North Dakota.-Local boards may furnish free text-books, and must on popular

vote.

South Dakota.-A county board of education is required to adopt a uniform series for five years, to be furnished through designated depositories under publishers' guaranty. On petition of a majority of electors a school corporation must arrange for free text-books.

Vermont.-County authority fixes a list for five years on publishers' guaranty. On popular vote local boards furnish free text-books.

Wisconsin.-District board fixes list for three years. Penalty, on every member of the board, $50. On popular vote books are furnished free without time limitation as to change.

THE LOCAL UNIT OF SCHOOL ORGANIZATION.

The State exercises remote authority over all public schools in its borders; the county, in most States, has a closer supervision of all schools in its limits, but has very little to do with schools in New England. In certain States it becomes the unit for the entire local administration of public schools. The town or township takes more or less of the local functions in other States, and the district becomes a local unit for variable functions in yet others. In 35 counties of Texas is a community system. The summarized statement below shows the principal agency through which local support and control of schools is exercised, special laws excepted, under which cities, towns, and independent districts exist. Counties generally receive, hold, and disburse moneys for townships and districts formed by subdivision of counties; towns or townships generally hold the same relation to districts formed by division of town or townships; in a few States districts have their own tax collectors and treasurers.

County.-Alabama is all either town or township. Florida, with provision for districts of united power; Georgia, Louisiana, recognizing Congressional townships in accounts of 16th-section land funds; Maryland, Mississippi, with provision for separate districts; North Carolina, with districts capable of holding real estate; Tennessee, with some local functions in districts, and only supervisory powers in subdistricts; Utah, with provision for division.

Town or township.-Alabama, the Congressional township for administrative convenience, its officers appointed and its accounts kept by county officers; Connecticut. the town may abolish districts; Illinois, township based on Congressional township, or district optional; Indiana, New Jersey, Ohio, and Pennsylvania, each township, incorporated town or city (or borough in Pennsylvania), district corporation for school purposes; Iowa, township based on Congressional township with subdistricts for supervisory convenience and independent districts, both in use; Maine, Massachusetts, Minnesota, township may be a district as a part of a county; New Hampshire, New York, recognized for certain land funds, but districts, generally; North Dakota, based on Congressional township; Rhode Island, may create or abolish districts; South Dakota, based on Congressional township; Vermont, Wisconsin, optional in formation of districts.

District.-Arkansas, Arizona, California, Colorado, Connecticut, where not abolished by the town; Delaware, Florida, Idaho, Illinois, optional with townships: Iowa, independent districts as well as townships; Kansas, Minnesota, Missouri, districts may be less than townships; Kentucky, Michigan, Mississippi, optional; Montana, Nebraska, Nevada, each village, town, or city is a district; New Mexico, New York, commissioners' district, a county or part of a county, has supervisory authority, school districts are parts of commissioners' districts, towns recognized for certain land funds; North Carolina, Oklahoma, Oregon, South Carolina, Tennessee, with limited powers, as stated under county: Texas, but cities may acquire exclusive control over their schools, towns and villages may be incorporated for school purposes only; in 35 community counties families associate from year to year to support schools and draw their share of public money; Utah, permissible as stated under county: Virginia, West Virginia, corresponding geographically to magisterial districts; Washington, each city or town (incorporated); Wisconsin, optional (see town or township); Wyoming.

CHAPTER XXXIV.

THE GROWTH OF TORONTO CHILDREN.

By FRANZ BOAS.

INTRODUCTION.

In 1891, when active preparations for the World's Columbian Exposition were being made, Prof. F. W. Putnam, curator of the Peabody Museum of American Archaeology and Ethnography, and chief of the Department of Anthropology of the Exposition, placed me in charge of the section of Physical Anthropology. At an early time during the preparation of the exhibits we agreed upon a plan to represent as fully as possible the growth and the development of American children. Valuable material was available, but it seemed desirable to extend the investigations over regions in which heretofore no observations had been collected. I submitted our plans to Mr. James Hughes, superintendent of public schools in Toronto, Ontario, and to Prof. Earl Barnes, of Leland Stanford Junior University. Through the interest taken by these gentlemen I have been enabled to obtain series of measurements of the school children of Toronto and of Oakland, Cal. The former series was taken under the supervision of Dr. Alexander F. Chamberlain, the latter under the direction of Prof. Earl Barnes. In both of these series the same plan, excepting details, was followed.

In the following pages I intend to present some of the results of our investigations upon the children of Toronto, together with a few general statements which the treatment of the material and a study of the questions involved suggest.

Since in previous investigations the influence of sex, of occupation of parents, and of descent had been studied, it seemed desirable to investigate the influence of other causes. I selected for this purpose the influence of the order of birth, i. e., the question whether first-born children have a development different from that of later-born children.

The blank on which the observations were recorded was drawn up so as to cover not only this point, but also the influence of nationality and occupation of the parents-facts which it was necessary to note in order to determine their influence on the questions to be investigated. Blue blanks were used for females, white blanks for males. Here is a copy of both sides of the blank:

[Front.] FEMALE.

No.

Record all linear measurements at nearest centimeter; all weights at nearest pound.

(name in full).

Observer (name in full).

Place of observation.

School.

[Back.]

EXPLANATION OF MEASUREMENTS.

1. Height standing.-Let the person stand close to the wall in front of the measuring rod. His heels must be close together, touching the wall, and he must stand perfectly straight, looking straight ahead without raising or dropping the chin, the head touching the wall. Read off the height of the crown of the head by means of the triangle, pressing one side against the rod, the other against the crown of the head.

2. Height sitting.-Put a low, level seat (for instance, a small wooden box) in front of the measuring rod. Let the person sit on it so that his knees are about 5 inches higher than the seat, which is accomplished by making the seat sufficiently low, or by using a footstool. Let the person sit far back, close to the wall, keeping his back erect against the wall. He must look straight ahead without raising or dropping the chin, the head touching the wall. Give the heights of the seat and of the crown of the head.

3. Finger-reach. Let the person touch a vertical post or wall with the second finger of one hand, and stretch along the measuring rod as far as he can reach. The rod must be held horizontally at right angles to the wall, in front of the body, along the median line of the arms. Let the person make the greatest possible efforts in stretching out his arms before you read off the figures.

4. The weight is to be taken in ordinary indoor costume.

The instrument used was a rod divided into 210 centimeters. The index arm was a separate piece, consisting of a wooden angle, the sides of which were long enough to insure perfect contact with the measuring rod and with the crown of the head of the person being measured. The head measurements were taken by carefully trained observers; the others by the teachers. The personal data were given by the parents of the children. Owing to the peculiar social conditions prevailing in Toronto, certain groups of the population are represented by very few individuals. This is true particularly in regard to the French population, the greater part of whose children seem to attend the parochial schools. There are also very few children of the professional classes included in our material.

The success of the collection of measurements is largely due to the assistance extended by Mr. James Hughes, and to the lively interest on the part of the teachers who undertook the arduous task of obtaining the necessary information from the parents, and who took many of the measurements. To all of them my thanks are due. I have also to thank Dr. Alexander F. Chamberlain for the efficient management of the whole undertaking.

THE METHODS OF TREATING STATISTICS OF GROWTH.

The treatment of anthropometrical observations, particularly of growing children, offers peculiar difficulties. During the past years a vast number of observations referring to the growth of children have been accumulated. The method of treating the results of such observations has largely been a comparison of averages and of the frequency of occurrence of measurements between certain limits; for instance, frequency of occurrence of statures from inch to inch or of weights from pound to pound. It is generally assumed that these figures express immediately the physiological facts relating to growth.

In almost all cases the observations have been taken only once, and on a great number of individuals, not repeatedly through a long number of years on the same individuals. For this reason the series, when arranged according to years, will not be homogeneous. The younger groups contain many individuals who will not reach the adult stage, while the older classes contain only few individuals who will die before becoming adults. When we assume the whole series to be homogeneous, we imply that the value of the measurement under consideration has no relation to the liability to die at a certain age, which assumption seems to be very doubtful. Without considering details, it would seem very likely that individuals far remote from the average, who show either too small or too large measurements, approach the limits between pathological and physiological variation, and are therefore more likely to die. This would imply a greater variability of the measurements of deceased individuals of a certain age than of living individuals of the same age. The series of living individuals of all ages can be equally constituted only when the measurements of the living and of the deceased show the same values. This fact has already been pointed out by H. Westergaard.

There are a few series of observations which seem to make the identity of the series of measurements of the living and of the deceased individuals of the same age very improbable. The most important among these is the peculiar decrease in the brain weight in males after the twentieth year. This can hardly be explained in any other way than by assuming an increased death-rate among men with very large brains at an age of about 20 years.

1 A centimeter is nearly 0.4 inch.

2 Grundzüge der Theorie der Statistik, p. 188.

Bowditch and Roberts have shown that, on the average, children of well-to-do parents are taller and heavier than those of poorer parents. Carlier has shown the same phenomenon by proving that a number of children of a certain class, when brought under more favorable conditions-in his case into a military training school-grow more rapidly than the rest, who were left in their former conditions. The mortality of children is greater among the poorer classes than among the well-to-do classes. Therefore among the young children a greater percentage of the individuals measured belongs to the poorer classes, whose children are at the same time shorter of stature than among the older children. This fact affects undoubtedly the averages of measurements collected in our public schools.

It does not seem unlikely that the correlation between measurements and mortality is more strongly emphasized at certain periods than at others. If, for instance, many individuals. of retarded growth should die during the period of adolescence, this might give the real explanation of the curious overlapping of the curves of growth of boys and girls, the girls between about the twelfth and fourteenth years being heavier and taller than boys of the same age. I am strengthened in this opinion by the observation, made by Dr. G. M. West, that the extent of this period and the amount of overlapping are the smaller the more favorable the conditions under which the individuals live. It would be interesting in this connection to study the curves of growth of a people which has a very high death-rate among young children.'

Social causes are apt to introduce other complications which restrict the comparability of the results. The poorer classes do not send their children to school as long and as regularly as the well-to-do; consequently their proportionate number among the school children decreases steadily, and this changing composition of the series must affect the results of the measurements.

The state of health of the children also affects the series, particularly during the first and last years. Weak children will be sent to school later than strong children, and the teachers will be inclined not to promote them as rapidly as strong children, so that the oldest school children will include an undue proportion, not only of those who are dull, but also of those who have been of weak health during a portion of their life.

For all these reasons, investigations based on single observations of children of various ages do not give us results which can be considered to indicate with the highest accuracy attainable the processes of human growth. The series for the various years differ in composition, and the physiological constants are therefore modified to a greater or less extent by a variety of disturbing factors. In order to obtain the physiological results with the greatest accuracy, the material upon which we base our studies must be made homogeneous. This can be accomplished in two ways. A very large number of children may be measured once; and year after year those who die and those who on account of social reasons are removed from the field of observation must be eliminated from the list. When all have become adults, the remaining individuals and those who dropped out for various reasons must be treated separately. But the best way would be to take measurements of a large series of children at stated intervals, as in this manner the fullest information on the manner of growth will be given, and as these repeated measurements furnish all the necessary material for subdividing the series so that each division will be homogeneous.

These limitations must be borne in mind in interpreting results of a single set of observations on children of various ages, or, to use Hertel's term, in interpreting results obtained by the generalizing method.

Besides this, certain corrections must be made which heretofore have not received sufficient attention. The number of children of various ages who have been measured is not equal. All the series begin with comparatively few children. The number increases from year to year until, beginning with the tenth or eleventh year, it decreases again. The change of numbers is not equal in the two sexes. It follows, from this fact, that among 6-year-old children, for instance, there are in the measured series more of the age of 6 years and 11 months than of 6 years exactly; and that, on the other hand, among the 15-year-old children there are more of the age of 15 years exactly than of 15 years and 11 months. In treating the various series of observations, all children between 6 and 7 years, 7 and 8 years, etc., or all the children between 6 and 71 years, etc., have been grouped together and the series is assumed to represent the sizes for the average ages, i. e., 6.5 years, 7.5 years, etc., or, in the other case, 6, 7, 8 years. On account of the varying frequency of the several months this is not quite correct. Among the young children the average will be

1 I expressed these views first in Science, Vol. XX., p. 351, December 23, 1892.