472 PROCEEDINGS OF SECTION F. 



population, but they may be signiticant only of the existence of 

 a criminal element, which is like a disease from which a commu- 

 nity suffers, but a disease of a superficial, and not of a vital, 

 character. One population may thus have more crime in it 

 than anothei', even much more crime, but substantially the two 

 peoples may be almost alike, the extent of the criminality in 

 both being quite immaterial. Say, for instance, _ that the 

 criminal population by which almost all the crime is done in 

 one country is 1 in 500, or l-5th of 1 per cent., and in another 

 population it is 1 in 250, or 2-5ths of 1 per cent., is not the 

 criminal element in either so small as to tell you nothing of the 

 general constitution of the people ? Not only, therefore, must 

 criminal statistics be used with care as far as the mere data are 

 concerned, but the difficulty of using them as indicative of the 

 general qualities of a population is overwhelrning. They can 

 only be used, if used at all, in conjunction with much other 

 information and statistics. 



The statistics bearing on sexual morality are equally difficult 

 to handle. The test here that is most commonly used is that of 

 illegitimacy; but the truth is that illegitimacy by itself tells 

 little, for the simple reason that in a town community there 

 may be prostitution without illegitimate births, whereas in a 

 rural community there may be even 'ess profligacy than in the 

 town, but with a larger number of illegitimate births, in conse- 

 quence of there being no prostitution. In one country also the 

 births may be registered as legitimate, through the children 

 being born in wedlock ; but this may go along with a general 

 laxity of morals of a remarkable kind. Sexual immorality is 

 also like crime itself, even when it can be measured oa. the same 

 basis in two different communities, more or less a thing apart, 

 and it may or may not be significant of the general morale of 

 the population. I suppose it is true, for instance, that the 

 rural population of Ireland stands bettei', as far as statistics of 

 illegitimacy are concerned, than that of Scotland, but it would 

 be a rash inference that in general morale the rural population 

 of Ireland is superior to the Scotch. For certain purj^oses the 

 statistics are good enough, but they must not be pushed to con- 

 clusions they do not bear. 



Statistics as to drunkenness also require a good deal _ of 

 careful handling. In fact, I see no way myself of establishing 

 statistically that one population is more or less drunken 

 than another. Apart from the difficulty already re- 

 ferred to, arising from the different distribution of 

 two populations according to age, so that one population has 

 proportionately more adults than another, and consequently 

 has a larger proportion of convictions for drunkenness and a 

 larger proportionate consumption of alcoholic liquors, — the two 

 tests usually applied in such comparisons, — it has to be con- 

 sidered that the tests themselves are not very good. The con- 

 victions for drunkenness, it is plain, like convictions for crime 

 generally, may be very largely a matter of definition and of 

 police administration. Before comparisons can be made the 

 state of legislation and of police administration in the countries 

 compared must be considered. As x^egards the consumption of 



