INFLUENCE OF ORDER OF BIRTH, ETC. 303 



than among the affected persons themselves. This would be the 

 case if the affected families were small. It is very desirable to 

 have data on the relative position of the affected person in indi- 

 vidual families of two, three, four, five, etc., persons so that it 

 could be ascertained whether or not within the limits of families 

 of a given size the marked individuals occur in preponderating 

 numbers in any given position. Data grouped in this way would 

 enable us to avoid several pitfalls incident upon handling mass 

 statistics. In the data of Weeks on the order of birth of epileptics 

 there is, as Pearson states, "no excess of the eldest-born in the 

 individual families; if there be any excess it is in the interme- 

 diates. Thus, if we may trust this data, which are slender, there 

 is no weighting of the first-born in the case of epilepsy unless it 

 arises from the weighting of small families." Treating the data 

 by the methods employed in other cases Pearson finds an excess 

 of epileptics among the first born. "We must, I think, conclude," 

 he remarks, "by recognizing that, while there is a weighting in 

 epilepsy, this is due to a selection of families rather than to a 

 selection of the elder-born in each family." How far the rela- 

 tively large proportion of first-borns in Pearson's data on other 

 defects may be due to the selection of small families is, of course, 

 uncertain. It is of value to know, however, whether the relative 

 preponderance of the first born in pathological stocks is due to the 

 smallness of the family. As Pearson remarks, "We are shooting, 

 so to speak, at the entire population of first boms, and a bias with 

 regard to selection of weaker families may come in, in much the 

 same way as families up to six or seven may be the sign of healthy 

 parents, and so the offspring will be less liable to disease. This 

 idea cannot be excluded. But in itself it indicates how inadequate 

 is the proposal to treat the problem only within families of con- 

 stant size." 



However it happens that the first bom in the population in 

 general comes to be selected for defect or disease, the reduction 

 of the size of families leading to an increase in the relative propor- 

 tion of first born individuals will inevitably cause an exaggeration 

 of several undesirable hereditary traits. In so far as the birth 



