22 



schools of the same class in towns, 54|- ; in the upper middle class 

 schools, 54 inches ; the lower middle class schools, 53 J ; in 

 elementary schools among agricultural labourers in the country, 

 53 inches ; among artizans in towns, 52-|- ; in factories and work- 

 shops in country districts, 52 ; in town districts, 51|^ ; in military 

 asylums, 51 ; in pauper schools, 50 J ; in industrial schools, 50 : — 

 the mean height of the whole male population at that age being 

 52| inches (4ft. 4|in.). The gradation is so regular and so uniform 

 that it would almost lead one to doubt the accuracy of the returns, 

 for every statistician Avill know that you must expect irregularities 

 in any curve derived from observation, which you correct by a 

 mathematical process ; but here these are the results directly 

 derived from observation, though almost too good to be true ; and 

 they show clearly the degradation of stature as boys are removed 

 further and further from the most favourable conditions of growth. 

 The same state of things will be found to exist among adults. 



Upon the interesting question as to the effect of heredity 

 upon height and other physical characteristics, Mr. Galton has 

 made some curious observations. The idea occurred to him about 

 ten years ago of testing it by experiments on the produce of seeds 

 of different size but of the same species. He derived from them 

 a law that offspring do not tend to resemble their parents, but 

 to be more mediocre than they. Anxious to confirm or dis- 

 prove his conclusion by anthropological evidence, he made an 

 offer of prizes for family records, carefully avoiding any allusion 

 to this particular purpose for which they were desired. The 

 result was even more striking than that of the experiment with the 

 seeds. He obtained the heights of 930 adult children and their 

 parentages 205 in number. In dealing with them he raised the 

 female heights to their corresponding male equivalents by multi- 

 plying them by 1-08. The height of the female parent thus 

 corrected and of the male parent together divided by two, give the 

 height of the mid-parent ; and the observations resulted in showing 

 that where the mid-parents are taller than mediocrity, taken as 

 5ft. 8|in., their children tend to be shorter than they by two-thirds 

 of the difference ; and the converse when the mid-parents are 

 shorter than mediocrity, when the children tend to be taller in the 



