20 Mr. Francis Galton on 



The intention is to start from the brood of a single 

 pair of moths (Selenia illustraria is suggested), and to 

 trace the changes of some one characteristic, — say of 

 the wing-length, — during many successive generations, 

 down three parallel but contrasted lines of descent ; the 

 broods being reared all along out of doors, and under 

 healthy and perfectly identical conditions. It will be 

 convenient to distinguish these three lines of descent 

 by the letters A, M, and Z. The A line of descent is to 

 be produced by selecting and mating one (or more) pair 

 of the longest-winged males and females that are found 

 in the brood of the original pair. The progeny of this 

 selected pair I will call A i. Out of the males and 

 females in A i. one (or more) pair of the longest-winged 

 will be selected. The progeny of this selected pair will 

 be A ii. Out of the males and females in A ii. one 

 (or more) pair of the longest-winged are again to be 

 selected ; their progeny will be A iii. This process will 

 be continued for, it is hoped, at least six generations, — 

 that is, for three years. Every moth in each of the 

 broods is to be preserved, and those which become the 

 parents of the succeeding brood are to be labelled 

 accordingly. It is intended to mount the moths in an 

 orderly series, separating the males from the females, 

 and severally arranging them in the order of their wing- 

 lengths, beginning with the longest and ending with the 

 shortest. 



The Z line of descent is to be produced by selecting 

 and mating one (or more) of the shortest- winged males 

 and females out of the original brood. Their progeny 

 will be Z i. Out of these a selection of one (or more) 

 pair of the shortest- winged males and females will be 

 made, whose progeny will be Z. ii., and so on as before. 

 Moths which are small, owing to deformity or obvious 

 unhealthiness, should not be selected for breeding from. 



The M line of descent is to be produced by selecting 

 and mating one (or more) pair of medium-winged in- 

 dividuals out of the original brood ; their progeny will 

 be M i. Out of these a selection of one (or more) pair 

 of the medium-winged males and females will be made, 

 whose progeny will be M. ii., and so on as before. It 

 will be understood that by medium-winged I mean with 

 reference to the brood in question, and not with reference 

 to the original brocd. The use of the M line is to 



