528 TRANSMISSION 



This series (|, J, J, ^g, etc.) carried to infinity would account 

 for the total heritage, and if these fractions are correctly taken 

 the influence of each separate ancestor would, on the average, 

 be represented by the fractions in the last column. 



This " law," at first somewhat arbitrarily derived, and an- 

 nounced with considerable hesitation, has received general sup- 

 port from later investigations, and all researches, mathematical 

 as well as otherwise, tend to establish its substantial accuracy. 

 The first announcement was based upon studies in stature. 

 Somewhat later opportunity was afforded to make exhaustive 

 studies from a large number of Basset hounds of distinct colors 

 and of several generations of known ancestry. This study is 

 reported in full by Galton, 1 and the results conform substantially 

 to the " law," which, as Galton observes, is " strictly consonant 

 with the observed binary subdivisions of the germ cells, and the 

 concomitant extrusion and loss of one half of the several contri- 

 butions from each of the two parents to the germ cell of the 

 offspring."' 2 



These Basset hounds were especially favorable for a study of 

 this kind. They are of two colors only, "lemon and white," or 

 'they may be marked in addition with a third color (black), in 

 which case they are known as tricolor. 



It is said that individuals are distinctly of one or the other 

 class, and that transitional specimens are very rare. The pedi- 

 grees and color descriptions had been carefully kept by Sir 

 Everett Millais, who had originated the particular stock. 



Some 8i6 3 hounds of known color were descended from 

 parents of known color ; in 567 3 cases the colors of the grand- 

 parents also were known, and in 188 cases the colors were known 

 for three generations back. Assigning fractional values to 

 parents, grandparents, etc., according to Galton's law of an- 

 cestral heredity, and calculating what the descendants should \>Q 

 under the law, it appeared that, according to theory, there should 

 have been 180 tricolor hounds descended from those whose 

 ancestors were known for three generations. In fact, 181 such 



1 Proceedings of the Royal Society, LXI, 401-412. 2 Ibid. p. 403. 



8 Not 817 and 577, as printed. There seems to be an error of i in the first row. 

 Proceedings of the Royal Society, LX, 409. 



