From A. R. Smith   16 October 1877

Dear Sir

In your “Descent of Man” (Chapter VIII “On the Relation between the period of Development” etc) you give as a general rule that characters inherited by the sex they first appeared in only appear late in life.¹ It struck me that the following might be one of the causes of the exceptions: that they first appeared late in life but have been transmitted at earlier & earlier periods, according to the rule you give in “Inheritance at corresponding periods of Life”, till at last they appear about the time of birth & that by the time they had worked down (so to speak) to this point they had become correlated with sex.²

perhaps exceptions to converse of the second rule, viz that characters which are inherited by both sexes generally appear early in life, are occasionally caused by a reverse process.

Hoping that you will excuse my writing to you, | I remain | Yours truly | A R Smith

P.S. In your preface you say that the causes of correlation are unknown.³ Would it not be sometimes caused (as I have taken for granted) by the correlated characters having appeared together for a long time & so, through inheritance, have acquired a strong tendency to appear together?