CONSUMPTION IN CANADA DAVIDSON. 17 



where also the conditions do not show violent extremes, are 

 the five towns selected from the Ontario Bureau of Statistics 

 Reports for detailed analysis. It is probable that the very large 

 percentage of large houses in St. John is an indication, not of a 

 large wealthy class, but of lack of prudence and foresight in the 

 inhabitants in the years which followed the great fire. We 

 might therefore conclude that in two towns only, Quebec and 

 Montreal, do the extremes of wealth and poverty show them- 

 selves ; and that the average condition is also the condition of 

 the great majority of the inhabitants of Canada. We might, 

 perhaps, also conclude that the average income obtained from the 

 returns made to the Ontario Bureau is not far below the average 



O 



income in Canada. Mr. Ames shows that in the district he has 

 investigated there is a weekly average income of $10.20 per 

 family, an average monthly rental of $8.73 per family, or 18 per 

 cent, of the family income, and an accommodation of 5.02 rooms 

 per family (U. S. Bulletin of Labor, p. 44 . The average weekly 

 income of the towns in Ontario is nearly $9.00, of which 17 per 

 cent, is expended on rent in places where rents must be much 

 lower than they are in industrial districts of Montreal, and where 

 accordingly better accommodation will be given for the money. 

 We may readily infer that the returns have been made by the 

 occupants of houses of 7 or 8 rooms ; and an overwhelming pro- 

 portion of the inhabitants of the towns, of which an analysis of 

 the house accommodation has been made above, occupy houses 

 containing from 5 to 10 rooms. Since, according to Engel, and 

 according also to the best canons of local taxation, the expendi- 

 ture on house rent is the best indication of income, we might be 



o 



safe in concluding that the average income set down above is the 



o o 



average for Canada ; but at the best the conclusion is problematic 

 and based on a series of assumptions and inferences from data 

 which are themselves only approximately correct. 



The main question is the actual accommodation obtained for 

 PKOC. & TRANS. N. S. INST. Sci., VOL. X. TRANS. B. 



