302 Saunders. 
(indicated by H, K and C, R respectively. The case of singleness 
constitutes only another example of a phenomenon already shown to 
occur in these same strains. In the case of sap-colour and of 
hoariness it is possible to put the supposition to actual proof. The 
C white and the R white can be bred together, and though indisting- 
uishable in appearance, the fact that they contain each the factor 
for sap-colour which is wanting in the other is proved by the fact 
that they yield a coloured F,. Similarly the glabrous sap-colour 
with H and the glabrous sap-colour with K can be detected by the 
fact that they yield a hoary F, when crossed together. 
Though the complete sterility of the doubles renders a similar 
test as to the nature of singleness impossible, the existence among 
these doubles of three types differing in constitution though alike in 
appearance is not inherently more improbable than the existence of 
the differently constituted whites, or of the differently constituted 
glabrous types, of both of which proof is obtainable. 
In regard to the second assumption (occurrence of partial coupling) 
we also find a parallel in the case of other characters in Stocks. The 
results obtained in regard to plastid colour and type of sap-colour 
(whether of a full or pale colour) show that the pair of allelomorphs 
determining plastid colour, viz. W (= white plastids) and w (=cream 
plastids) are partially coupled with the pair producing pale (P) or 
full (p) sap-colours. If two forms are mated together, having the 
constitution PW and pw respectively, then since white plastid colour 
is dominant to cream and pale sap-colours to full colours the simple 
Mendelian expectation in F, will be in the following proportion, — 
g pale sap-coloured with white plastids (PW) 
33 full 6 x 5 2, = (pW) 
3 pale ” ” ” ” 7 (Pw) 
I full ” ” ” ” ” (pw) 
But this result is not what occurs. The two most frequently occurring 
forms are those representing the combinations occurring in the parents 
— in this case PW and pw, the recombinations Pw and pW are com- 
paratively rare. What we find in fact is 
very many pale sap-coloured with white plastids 
few full a - A 5 > 
few pale 5 n » cream . 
many full „ re & - > 
The appearance of an unexpectedly large number of the form 
exhibiting both recessive characters becomes intelligible on the 
