HUMAN FACULTY   267

being equal to that of its most highly gifted representative at the present moment.

I desired to plan a laboratory in which Human Faculty might be measured so far as possible, and, after much inquiry and trouble, drew up and sent a printed circular to experts, showing in outline what seemed to me feasible, and drawing attention to desiderata. Useful replies reached me from many quarters.

There was no one to whose intelligent co-operation I then owed more than Professor Croom Robertson (1842-1892) of University College. His genius and temperament were of the most attractive Scottish type -exact, sane, and very genial. He was well known by his work on Hobbes, and as the founder and Editor of the periodical .Mind, in which his critical notices of current philosophical literature were soon recognised as of especial weight. He was a thorough friend, whose death left a void in my own life that has never been wholly filled.

The leading ideas of such a laboratory as I had in view, were that its measurements should effectually

sample " a man with reasonable completeness. I t should measure absolutely where it was possible, otherwise relatively among his class fellows, the quality of each selected faculty. The next step would be to estimate the combined effect of these separately measured faculties in any given proportion, and ultimately to ascertain the degree with which the measurement of sample faculties in youth justifies a prophecy of future success in life, using the word

success" in its most liberal meaning.

The method of centiles (or of per-centiles as I originally called it) was devised to give greater pre-