Il8 GREAT MEN OF ANCIENT TIMES. 



we have 129 years as the least reasonable length of time elapsing 

 between the birth of the founder of the family and that of Laurent 

 Pierre. Dividing this by three we get an average of 43 years, 

 which would give us three successive generations born in our 

 class B. Applying these figures we find that the first unrecorded 

 person was probably born about 1706, or later than Bernard, and 

 the second unrecorded person probably about 1749. Inserting these 

 dates in the diagram we find that they accord well with the known 

 date of birth of Antoine Laurent. We know that Adrian, the son 

 of Antoine Laurent, is in class A, because his father was 49 when 

 he was born. We thus have in the de Jussieu family four births 

 in class B and one in class A. Under such circumstances I have 

 not hesitated to rank these persons in this way because I know 

 that they cannot rank lower, while there is a possibility and even 

 a probability that some of them rank higher. To bring these per- 

 sons as low as class C would be to make Antoine born when the 

 founder of the family was 13 years old or less, an assumption that 

 is not only unreasonable but absurd. 



RELATIVE BIRTH-RANKS IN LARGE FAMILIES. 



In looking through biographical sketches, when no definite dates 

 are given, we have to seize on such expressions as "eldest son," 

 "youngest son," "third child," etc. Unaccompanied by information 

 as to the size of the family, such expressions tell us very little, 

 but when they are accompanied by the statement that the family 

 consisted of 10, 12 or 15 persons, then they are very persuasive. 

 To determine just what such expressions mean I had recourse 

 again to the Redfield Genealogy. By tabulating a large number of 

 families I found that in families of five, fifty per cent of the youngest 

 children were equally divided between classes A and B ; in families 



