Correlation and Application of Statistics to Problems of Heredity 137 



irregularity in the mode of life that was greater two or three generations 

 back than now. Further he found that the "prime" for weight was also 

 earlier in age for the older generations, being hardly discoverable at all in 

 those born in the first third of the nineteenth century or in the professional 

 classes of the 'eighties. His three smoothed curves reproduced on p. 136, 

 with the table of mean weights at each central age, indicate that noblemen 

 of the generation which flourished about the beginning of last century 

 attained their meridian and declined much earlier than those of the genera- 

 tion sixty years their juniors, or indeed than the mid- Victorian professional 

 classes, where the culminating point was difficult to ascertain. 



Galton's data were somewhat scanty as the following table will indicate, 

 but his general conclusions appear to be justified : 



Actual Mean Weights in pounds at Various Ages. 



"There can be no doubt," he writes, "that the dissolute life led by the upper classes about 

 the beginning of this century, which is so graphically described by Mr Trevelyan in his Life 

 of Fox, has left its mark on their age-weight traces. It would be most interesting to collate 

 these violent fluctuations with events in their medical histories; but, failing such information, 

 we can only speculate on them, much as Elaine did on the dints in the shield of Launcelot, 

 and on looking at some huge notch in the trace [for the individual], may hazard the guess, 

 'Ah, what a stroke of gout was there!' " 



Although no great importance can be attached to Galton's results for this 

 particular class of subject, yet the problems his paper suggpsts might be 

 profitably studied on more ample material now extant. I am therefore glad 

 to have brought to light once more this long forgotten paper. 



PGIII 



18 



