198 Prof. Forbes*s Experiments on the Weight, 



been ascertained before the commencement of the experiments, 

 anil was fonnd to be pretty constant throughout the scale. 

 But after the experiments were finished this was by no means 

 the case, the error having become variable owing to the inter- 

 fering action of a small spring employed to bring the index to 

 zero. As this, however, only aftects the absolute results (or, 

 at least, its relative influence is trifling), I have contented my- 

 self with applying an interpolated correction deduced from 

 the mean of the errors before and after, which cannot differ 

 much from the truth. But the instrumental errors are not 

 the only ones to be contended with. To avoid errors in the 

 use of the dynamometer requires vigilant superintendence on 

 the part of the observer; and as the first pull is generally 

 (though not always) greater than the second or third, this 

 also must be allowed for. I have invariably repeated the ex- 

 periment three times, and often much more frequently. When 

 extraordinary cases have occurred I have taken the precaution 

 of observing at distinct intervals of time. 



In ascertaining the mean results the following method has 

 been adopted: the natives of each country were separated, 

 and each class divided, according to age, into twelve sets, from 

 14? to 25, the greatest number being of the age of 18 years. 

 The mean weight, height and strength for each year was com- 

 puted, and the result projected upon ruled paper. Curves 

 were drawn through the points thus projected, in such a way 

 as to represent most satisfactorily the whole observations. 

 These curves, with the determining points, are now exhibited 

 to the Society. It is proper to add that the ages registered 

 being the ages at last birthday, the weight, &c., registered is 

 not that due to the age noted, but at a mean to an age half a 

 year later. Thus all the persons who were 20 last birthday 

 are between the ages of 20 and 21, or 20^ at a mean. This 

 has been attended to in making the projections. 



Besides the English, Scotch and Irish curves, I have exhi- 

 bited those of the Belgian development, from M. Quetelet's 

 experiments, reduced to English measures. The thickness 

 of the shoes not being included in these experiments, half 

 an inch (perhaps too little) has been added to make them 

 comparable with the others. It is important to add that M. 

 Quetelet's experiments here quoted, as well as my own, were 

 made upon persons in the higher ranks of life, in both cases, 

 in fact, upon persons having the benefit of academical instruc- 

 tion. 



The number of persons examined by me in the two winters 

 before stated was thus divided : Scotchmen, 523 ; Englishmen, 

 178; Irishmen, 72; from the Colonies, &c. 56', total, 82y. 



