22 Result of Crossing Japanese Waltzing tvith Albino Mice 
correlation already mentioned between purity of ancestry among the albinos 
employed in making the original cross and the darkness of the hybrids. 
The albinos of the group seem more obviously distinct from the rest than do 
the albinos among the offspring of hybrids, and the ancestral coi'relations are 
materially affected by their retention or omission. 
Retaining the albinos the correlation between whiteness of coat and purity of 
the albino grandparent is about — 01 5 (cf. Table XLV.); between whiteness of coat 
in coloured individuals alone and purity of albino grandparent the coefficient rises 
to about — 0'25. We have here a further proof of the extent to which the ancestry 
of the " recessive " albino affects the coloured descendants of the cross. The 
correlations with the albino parent are too small to be worth noticing until the 
probable error of the method employed for their determination can be given ; the 
data are given in Tables XLIX. and L., where it will be seen that nearly all the 
albinos used as parents were pm-c-bred, so that the determination of correlations 
is difficult. 
The hybrid parent appears to be more closely correlated with the young, both 
in colour and in amount of whiteness, than the albino, the correlation in both 
characters being about 0"2, the value found among offspring of hybrids paired 
together. 
Hybrids crossed with extracted Albinos. 
The object of this exjDeriment was to determine if the percentage of albinos 
produced by such crosses would be different from that produced by crosses between 
hybrids and pure albinos. The number of albinos obtained — 59 out of 120 — is 
of course an excellent experimental half; but the following considerations should 
be carefully weighed, (i) the albinos used were not all produced by pairing two 
hybrids ; in many cases (see Table D) they were produced by crossing hybrid and 
albino, (ii) the hybrids used were in some cases (6ri.,, Gu, <?i3, G^is, (t,?, (tu,, (?.,o) the 
offspring of waltzers and extracted whites, (iii) the character the percentage of 
which one is trying to diminish in the offspring is identical with that of one of 
the parents : in offspring of extracted hybrids one was able to diminish the 
number of albinos by diminishing the amount of albino ancestry of the extracted 
hybrids ; but in this experiment, while one does diminish the albino ancestry, one 
of the parents themselves is always albino ; so that the effect of the diminution of 
albino ancestry is rather swamped. 
If separate tables are made for, (i) hybrids crossed with albinos produced 
by the union of hybrids, (ii) hybrids crossed with albinos produced by hybrids 
paired with albinos, and (iii) hybrids from extracted white crossed with extracted 
albinos, the numbers become so small as to be useless to base arguments upon. 
The occurrence of pink eyes in seven of the young, or five per cent, of the 
number produced, distributed among five families (more than one-fifth of the whole 
number), is clear evidence that these extracted albinos have a power of hereditary 
