lIvTRODUCTORY. LXXXIII 



1. There is a perfect form or type of man, and the tendency of the r 



•ace is to 



uttain this type.' 



2 The order of growth is reouLir toward this type. 



V The variations from this type foUow a definite hiw, the hiw of accidental causes. 



4 The Hne formed by these variations, when arranged in groups, recedmg on 

 either si<h^ of their mean, is the curve well known to mathematicians as the bmomial : 

 it was first applied by Newton and Pascal to questions of astronomy and physics, but 

 it is apphcable to all the qualities of man which can be represented by numbers. 



5 The more numerous the data obtained by actual measurement, supposmg them 

 to be made with reasonable care and without bias,nhe more nearly accurate is the 

 mean result, and th e more closely does it corres pmj^mthjh at obtained by calculation. 



rTTu^.^rk treats only of man, the applicability of Qnetelefs theory to other subjects bas not been alluded 



,., but extensive invostigations'iu both the animal and vegetable kingdon.s seen, to corroborate rts correctness and 



"^'V^rjexceedinglv large nnn.ber of measures, weights, orother nun.erieal determinations of any constant mag- 

 niUulebe ■ren-suppo:ingnobias,or any cause of error acting preferably in any one dnecfon to ex,st-„ot only 

 Si K number of mall ^rors vastly exceed that of large ones, but the results will be found to group themselv 

 rbout the mean of the whole always according to one invariable law of uuxuber, and that the n,on, precsely, the g.eater 

 the otal number of determinations. « ' Rude and unskillful „,easureu,ents of any knul, accumulated ,n very 

 greaTnt^betare cou.petent to atiord precise n,ea.n results. The only conditions are the con.jnual anunus 'nensuj^ 

 tS'seuceof bias, the correctness of the scale with which the measures are comi«red^ and the assurance that we 

 have the entire range of error, at least in one direction, within the record."-Sir J. F. W. Hersciiei,. 



