STATISTICAL COMPARISONS. 46T 



of Russia, because in a given period it has annexed a larger 

 area and a larger population than Russia had done, the truth 

 being that the area annexed by either country in tha period in 

 question was largely desert, so that it hardly counted one way or 

 the other and that the populations annexed were of most various 

 quality. The point of real aggressiveness or not was studiously 

 overlooked in this ingenious statistical comparison. Constantljr 

 at home, also, there are continual discussions on the balance of 

 power, in which the numbers of the populations and the armies 

 they can put in the field are simply counted ; whereas the whole 

 question turns largely upon the quality of the respective 

 populations and the state of their warlike preparations, and not 

 so much ujjon mere numbers. The question of quality of 

 population arises in a difierent way in those political questions 

 which are settled by numbers at the ballot-box in democratic 

 communities, and I am not sure but that some of the underlying 

 assumptions of politics are based on the refusal to recognise 

 the essential difierences of difierent peoples, as, for instance, 

 in the concession at home to the people of Ireland of an 

 equality of voting power and representation in the Imperial 

 Parliament, and, i-eally, far more than an equality, whereas, in 

 some qualities, such as wealth, they cannot be regarded as equal, 

 although they may be equal, or superior, in yet other qualities. 

 Common-place, therefore, as it seems, to say that when we see 

 columns of comparative figures of population we must not 

 assume the units to be alike, the applications of the doctrine 

 are not really common-place. We are all subject to the influ- 

 ence of unexpressed and underlying assumptions, and I have 

 only given a few out of many possible illustrations of the 

 dangers that may arise in using these very ordinary figures 

 without constantly thinking of what they mean. 



I come finally to less debateable ground in one way, but where 

 there is practical mischief from the misuse of figures. Nothing 

 is more common than to compare populations which may be 

 assumed to be racially very nearly alike, or approximating in 

 certain qualities, but which really difter greatly from each 

 other in regard to the distribution of the population according 

 to age. France and Germany, for instance, are continually 

 spoken of as if the difference of their numbei's made a corre- 

 sponding diflerence in their force. In fact, the population of 

 Germany contains a much larger percentage of children than 

 that of France does, and the numbers of adults in the two 

 countries do not difTer so much in proportion as their total num- 

 bers do. 



To show what differences there may be in the relative pro- 

 portions according to ages in different communities, I have 

 brought together certain figures extracted from the last census, 

 in each case showing the total numbers, the total male popula- 

 tion, the males above the age of 20, and the males between 20 

 and 40, in France, Germany, and the United Kingdom, respec- 

 tively. (See Table A annexed.) From this it will be seen that 

 France, with a population of close on 38 millions, has 11,828,000 

 males above 20 ; and Germany, with a population of just under 47 

 millions inhabitants, or upwards of 20 per cent, more than that of 



