98 



PROF. ERNEST W. MACBRIDE, M.A., F.R.S.. ON 



to rue that there is probably a considerable element of truth in 

 it. My reasons for this will be given later on. 



The next considerable advance upon Darwin's position was 

 made by Darwin's brother-in-law, Sir Francis Galton, who 

 invented a means of measuring at the same time the number of 

 variations and their amount. He applied his method to the 

 measurement of variations in man, but it was applied to other 

 animals by Weldon, who was Professor first in University 

 College and afterwards at Oxford. An example will make this 

 plain. Suppose we are desirous of finding how much the 

 breadth of the carapace of a crab varies, and how many broad 

 and how many narrow crabs there are, it is obviously of no use 

 to measure the absolute values of the breadth of the carapaces of 

 various crabs, because crabs vary in size. If. however, we take 

 the length of the crab as a unit and express the breadth of the 

 carapace as a fraction of it, then the value of this fraction is high 

 for broad crabs and low for narrow crabs. If we now determine 

 the value of this fraction, for say 500 crabs, and sort the values 

 into groups, the members of which differ from one another by 

 less than a certain limit, then we have the means of drawing a 

 curve which will show us at once the range of variation and the 

 number of specimens showing any given extent of variation. If 

 we measure along a horizontal line lengths proportional to the 

 values of the fraction, and erect at the point corresponding to 

 each value a perpendicular proportional in length to the number 

 of specimens showing this value, then we get by joining together 

 the summits of these perpendiculars a curve. If we take a great 

 number of specimens the curve becomes more and more sym- 

 metrical, proving that there is a certain mean breadth of carapace 

 which the great majority of crabs show, and that as we recede 

 from this value we find fewer and fewer crab?, but that on the 

 whole there are as many with the fraction at a higher value than 

 the mean as there are with the fraction at a lower value than 

 the mean, and that extreme values either above or below the 

 mean are very rare. Exactly such a curve as this is got if we 

 record hits made by shooting: at a target — most will fall at a 

 certain distance from the bull's-eye. There will be a very few 

 bull's-eyes and a few outers. Hence this curve is called the 

 curve of error. There is a school of scientists headed by Prof. 

 Pearson who seem to think that by this method we have 

 penetrated the secrets of variation, that all these variations from 

 the mean are inheritable, and that if natural selection were to 

 favour a greater breadth of carapace than the mean the deviations 

 necessary are. present in sufficient numbers, and in any event 

 only very few crabs of any generation survive. 



