470 PROCEEDINGS OF SECTION F. 



of causes themselves, one doctor giving a proximate and 

 another an ultimate cause. Before statistical comparisons canbe 

 made, something must be ascertained as to whether definitions 

 and method of registration are substantially the same in the two 

 countries compared. In historical investigations, even in the 

 same country, the precaution is equally indispensable. 



STATISTICS AS TO CHARACTER OF POPULATION. 



I proceed next to statistics, from which inferences are 

 commonly drawn as to the qualities of a population - I mean 

 statistics on such subjects as education, crime, sexual morality, 

 drunkenness, insolvency, and thrift. Onallthesepoints different 

 countries have statistics, which may have a meaning when they 

 are properly used, but which it is most difficult to use properly. 



To begin with education. Which is the most fortunate popu- 

 lation of the world as regards the general education of the 

 people? One often hears of the United States in this connection — 

 of the numbers of children of school age and the numbers attending 

 school as compared with less fortunate populations. But let me 

 take the following passage from a mem.orandum by Mr. Fitch, 

 one of Her Majesty's chief inspectors of training colleges, on the 

 working of the Free School system in the United States, France, 

 and Belgium : — 



" In England and Wales the calculations of average attendance 

 are made on the assumption that every school is open at least 

 400 times or 200 days in the year. It is on this basis that the 

 annual returns in the official report of the Education Depart- 

 ment state the average attendance of scholars in infant schools 

 and departments to be 68 per cent., and that in schools for 

 older children to be 82'2 per cent. But in the United States 

 there is no uniform or generally accepted rule respecting the 

 length of the school year. In the principal cities, especially in 

 the East and West, the schools are open 10 months out of 12, 

 and in these the statistics of attendance may be fairly compared 

 with our own. But taking the country through, the average 

 number of days in which the public schools are open is 129 in 

 the year, and this fact implies that in the country places, 

 especially in the South Atlantic and South Central States, the 

 number of school days falls much below that average. In 

 Alabama and in Georgia the schools are open only three months 

 in the year, the teachers are paid by the month, and hold no 

 permanent appointment. In Louisiana and Missouri the 

 small sum appropriated to education by the State barely 

 suffices to keep the schools at work more than four months in 

 the year. In Nebraska the returns for 5,407 schools show^ 3,904 

 to be kept open for six months and upwards, 529 for more than 

 four but less than six months, and 974 for less than four 

 months. In New Hampshire the average length of the school 

 term is 22"9 weeks ; in North Carolina it is 12 weeks ; in South 

 Carolina, 3^ months. In Texas the towns give an average of 

 eight months and the country districts five months. On the 

 other hand, in some of the h tlantic States the rate is much 

 higher. In Pennsylvania, exclusive of Philadelphia, in which 



