^72 lAfe aiul Letters of Franeis Galton 



did not Bee why a similar periodic anthropometric examination should not bo 

 made in all the gi'eat schools of the coxintry; and he makes the excellent 

 suggestion of an itinemnt anthropometric laiwratory, which should circulat( 

 from school to 8chi>ol. The idea is a very good one, and would keep t\\v 

 school laboratory, where such existed, up to modern standards of methods, 

 tests and a]>|)aratu8. 



DixcussuiiT what should be measured, Galton observes: 



"One object is to ascertAin what inaj' be called the personal constants of mature life. This 

 I ':• i-ie must not be taken in too strict a sense, because there is nothing absolutely constant in a 

  iUij iKxly. Life is a condition of jn-rpetuul change. Men arc alxjut half an incli shorter when 

 they go to bed than when they rise in the morning. Their weight is affected by diet and hnbit 

 of life. All our so-callinJ personal constants are really variables, though a large proportion of 

 their actual variations may lie between narrow limits. Our tirst rule then i.s, that the trouble 

 of measurement is best repaid when it is directed upon the least variable faculties." (p. 207.) 



He then touches on a point which still troubles the anthroporaetrician, 

 especially in the case of mental tests: 



"There are many faculties that may be said to be potentially constant in adults though they 

 •re not develojjed, owing to want of exercise. After adequate pmctice, a limit of etticiency 

 would in each case be attained and this would be the personal constant; init it is obviously 

 ini|M>.ssible to guess what that consUint would be from the results of a single trial. No test 

 professes to do more than show the efficiency of the faculty at the time it was applied, and nian\ 

 tests do even less than this, being so novel to the person e.xperimented on that he is maladroit, 

 and fails to do himself justice' ; consequently the results of earlier trials with ill-devised test-^ 

 may differ considerably from those of later ones. The second rule then is that the action- 

 required by the tests should be as familiar as possible'." 



Galton makes a suggestion as to practice which might be worth following 



up, although great care would have to be taken in experimenting between 



sessional and secular variations. The sessional variation, or variation in a 



sitting (or in one closely contracted series of tests), may be largely a result of 



chance, j)artly of practice, and partly of fatigue, while the secular variation 



may show the marked effect of continual practice\ The suggestion runs: 



"There is aoiue hope that we may in time learn to eliminate the effect of an unknown amount 

 of previous practice by three or more distinct sets of trials. There exists a rough relation lietween 

 practice and proficiency which ought to be apjMirent wherever progress is not due to acquiring 

 a succession of new knacks, but proceeds regularly. When no practice has previously taken 

 place, the progressive improvement will Ije very rapid ; then its rate will smoothly d«>crt'4ise until 

 it comes to an entire stop. I suspect that a curve might be drawn betwe<^n proficiency and 

 practice, and that the data afforded by at least three successive series of tests would roughly 



' I think Galton has conceivably overlooked a point here. One test of fitness to environ- 

 ment is the readiness with which an individual can adapt himself to new conditions and respond 

 promptly to the everchanging experiences of his life. In many cases therefore a iwvd test is a 

 truer tost of mental agility, than one which has liecome a familiar routina 



' This nde appears to lie stated too generally; when familiarity comes in, then there will be 

 a correlation l)etween length of practiw and efficiency, of which we do not know the intensity 

 When I was a boy I was taught to dovetail and learnt to do so creditably; any mediocre car- 

 penter would excel my work now Iwth in sjkhxI and neatness. On the other hand I should 

 proK. out far more clearly and rapidly than he would a page of letterpress. Any test 



to <\< 11 r n-lativo powers of mentally controlling the rapidly moving hand ought, I think, 



to l*j made in a manner which should be as unfamiliar to us lx)th as possible. 



• The rcatler may consult a paper by E. S. Pcui-koh on the " Variations in Personal lixjuation, 

 etc.," fiiometrxka, Vol. XIV, pp. 23-102. 



