COMPLICATIONS ARISING IN THE CROSS-BREEDING OF STOCKS, 145 
combinations the two factors meet and produce hoariness. For, in 
whatever way we pair forms such as white and cream and red, the 
offspring in F’, are in every case hoary. We are driven to conclude that 
in some way the surface character is dependent upon the flower colour, 
and, in fact, it is only when both these characters are considered in 
conjunction, that the system of inter-relationships underlying the above 
results becomes intelligible. 
The proportion of 9 sap-coloured to 7 non-sap-coloured in F, 
from white x cream, and of 8 sap-coloured to 1 non-sap-coloured 
in F, from white or cream xX a sap-colour is at once explained if we 
assume that two factors are necessary for the production of sap-colour, 
and that one of these is present in white and the other in cream. Red or 
any other sap-colour, ex hypothest, contains both. If we indicate these 
factors by C and R, and their absence by c and r, we may write Cr and 
cR respectively for white and cream, CR for any sap-colour. From the 
cross Cr x cR, where only individuals containing both C and R have 
coloured sap, we should expect in I, 9 coloured to 7 uncoloured ; 
from the cross CR x cR or CR x Cr we should, on the other hand, 
expect 3 coloured to 1 uncoloured in F,. ‘This, as previously stated, 
is precisely what occurs; by the above supposition we can therefore 
satisfactorily explain the general facts as regards sap-colour. We may 
now return to the consideration of surface character. Though the facts given 
under 2 and 8 point to the existence of two complementary factors which 
on meeting produce hoariness, the supposition that only one or other is 
present in each glabrous strain will not account for the results observed. 
A noticeable feature in both cases is the absence of non-sap-coloured 
hoary plants in F,. This fact furnishes the required clue. Hoariness, 
it appears, cannot be manifested unless C and R are both present together 
with H and K, 
If this be so, it follows that sap-coloured glabrous types must lack one 
or both of these factors; but on this view non-sap-coloured forms may 
contain both, or, again, may lack one or both. Hence if we denote the 
presence of these factors by H and K, their absence by h and k, glabrous 
sap-colours must be either CRHk* or CRhk in composition, non-sap- 
colours Cr (or cR) HK, Cr (or cR) Hk, or Cr (or cR) hk. On working out 
the results of these several alternatives it will be found that the expres- 
sions CrHK and cRHK for the two non-sap-colours, and CRHk for all 
glabrous sap-colours alone yield the results given above. The accom- 
panying figures will serve to make this clear. 
The foregoing scheme provides an explanation of the complications 
underlying the results obtained with regard to surface character and 
presence or absence of sap-colour. The diversity of F, forms appearing 
in certain crosses in the sap-coloured class are evidently due to other 
additional factors, which are merely superposed upon those already con- 
sidered, and call for little remark. Thus in F, from white glabrous x cream 
glabrous we can distinguish by inspection, without further breeding, 
ten distinct forms, viz. four self sap-colours— purple, red, plum, copper— 
* We are unable to distinguish between H and K; hence the choice of H rather 
than K in these expressions is purely arbitrary. The same applies to C and R in the 
case of white and cream. 
K 
