INFLUENCE OF MARRIAGE ON DEATH-RATE. 241 



matrimony is favourable to longevity. Yet the con- 

 sideration of a few parallel cases will suffice to show 

 how small a foundation the figures I have quoted 

 supply for such a conclusion. What would be thought, 

 for example, of any of the following inferences? 

 Among hot-house plants there is. observed a greater 

 variety and brilliance of colour than among those which 

 are kept in the open air; therefore the housing of plants 

 conduces to the splendour of their colouring. Or 

 again : The average height of Life Guardsmen is 

 greater than that of the rest of the male population ; 

 therefore to be a Life Guardsman conduces to tallness 

 of stature. Or to take an example still more closely 

 illustrative of Dr. Stark's reasoning: The average 

 longevity of noblemen exceeds that of untitled persons ; 

 therefore to have a title is conducive to longevity; or 

 borrowing his words, to remain without a title ' is more 

 destructive to life than the most unwholesome trades, 

 or than residence in an unwholesome house or district, 

 where there has never been the most distant attempt 

 at sanitary improvement of any kind.' 



We know that the inference is absurd in each of the 

 above instances, and we are able at once to show where 

 the flaw in the reasoning lies. We know that splendid 

 flowers are more commonly selected for housing, and 

 that Life Guardsmen are chosen for their tallness, so 

 that we are prevented from falling into the mistake of 

 ascribing splendour of colour in the one instance, or 

 tallness in the other, to the influence of causes which 

 have nothing whatever to do with those attributes; 



R 



