HYBRIDS AMONG THE AMARYLLL® AND CACTACER. 413 
(hort.) ]. Perhaps this instance may be claimed as partially confirmatory 
of Mendelism. 
(2) Primula kewensis [P. floribunda x P. verticiliata], see “ Gard. 
Chron.” 3/3/00; and 31/3/00 with fig. “Garden,” 2/06, p. 67; and 
8/06, p. 62. A chance hybrid of which only one plant was noted, but 
since verified by experiment at Kew [ Garden,” 3/2/06]. This is ap- 
parently an instance of a new species originating through a single 
hybridisation, for this hybrid reproduces itself from seed with the fixity 
of character usually associated with old-established species. It is nearly 
equipoised between either parent, but perhaps leans towards floribunda, 
which it resembles in its extended period of blooming, but shows some 
trace of the mealiness of verticillata in its foliage. In this hybrid we 
have an interesting instance of mutation in primary sexual characters. 
For some years the hybrid was regarded as sterile, but presently one 
plant was noticed to bear both the typical pin-eyed flowers and also 
some which were thought to be thrum-eyed. On pollinating the latter 
with the former, fertile seeds were, for the first time, produced. Since 
then the seedling plants usually carry both (alleged) classes of flowers 
and seed freely enough. Mr. Bateson maintains that none of the flowers 
are strictly thrum-eyed, but that the difference is merely one of length 
in the stamens. At least this is undoubted, that this mutation in a 
primary sexual character was associated with fertility in a plant previously 
sterile. This hybrid is not hardy, and does not thrive very well with me. 
Colour Changes. 
I have had no experience of new colours arising except through 
hybridisation ; other instances of colour-mutation which have come under 
my observation may be classed as reversion, recurring erraticism, or 
albinism. 
A Cineraria, crimson with white eye, isolated and self-fertilised, 
produced : 
19 crimson with white eye 
8 white self 
9 purple self 
1 pink self 
7 bicoloured (all blue and white) 
1 tricoloured (white, red, and purple) 
45 plants. 
The only apparently new colour is blue, appearing in 37 per cent 
as blue or purple. This is the ancestral purple of C. cruenta, appearing 
as red or blue, or (together) as purple. We find these three colours in 
some shades and combinations in 82 per cent., and the remaining 18 per 
cent. are albinos. No new colour appeared. 42 per cent. were the colour 
of parent, in 78 per cent. such colours were partially present, 40 per cent. 
showed some loss of colour, and 20 per cent. complete loss. 
A self-fertilised Gloxinia showed 65 per cent. the colour of the parent, 
100 per cent. showed some of the colour of the parent, 35 per cent. some 
loss of colour. No new colour appeared, 
