386 Wisconsin Academy of Sciences, Arts, and Letters, 



resulting error in a canclusion based upon a comparison of rec- 

 ords is evident, and I have tried to avoid it bj substituting for 

 a comparison between the records of successive periods in the 

 same community that between successive generations of the same 

 family or families ; but I have not as yet been able to make my 

 investigations sufficiently extensive to give a basis for anything 

 more than a mere suggestion. 'My w^rk has covered seven gen- 

 erations of three distinct families extending back to the middle 

 of the seventeenth century in Xew England, and now widely 

 scattered from their early home. The records of the two earliest 

 generations, however, were obviously so deficient in the dates of 

 death of children dying young that it seemied useless to consider 

 them, and so many of the seventh generation still survive that 

 that too, was necessarilv omitted. I have tabulated below the 

 statistics of the ages at death of the members of four generations, 

 divided into five classes with respect to age at death, giving the 

 per cent, of the whole dying within the limits of age of each 

 class 'and their mean age at death. 



The near approach to uniformity in the later columns of this 

 table evidently suggests a serious doubt of the correctness of the 

 theory of decreasing vigor as the generations pass, as far as so 

 limited an investigation can have any weight in dealing with 

 the question. 



