January, 191 | 
THE ORCHID WORLD. 
85 
most cases they must of necessity be 
erroneous. The smiple fact of the segrega- 
tion of characters m the second generation 
altogether upsets such a calculation. 
So far, we have dealt with the three par- 
ticular problems by applying the principles of 
genetics in a general way to enable the 
breeder to secure a definite result m the 
shortest possible time. Space will not allow 
one to deal with the interesting mass of minor 
details concerned in these three problems. 
One or two special points, however, may be 
noted. Judging from the Fi results recently 
produced, it would appear that in order to get 
a true Scarlet Crispum, Cochlioda Noetzliana 
should be crossed with an Odontoglossum 
with a white ground colour, e.g., O: crispum 
or O. nobile, and in order to avoid, if possible, 
the presence of purple sap m the segments it 
would seem advisable to use the albino O. 
crispum xanthotes and O.nobile albuir. in pre- 
ference to the tinged, blotched or plain forms. 
In order to get a Crimson Crispum it would 
appear that Cochlioda Noetzliana should be 
crossed with an Odontoglossum with a yellow 
ground, e.g., O. triumphans and O. luteopur- 
pureum, and to avoid the dull purple tint in 
the crimson, the xanthic albinos O.-t. Ajax and 
O.-l. Vuylstekeanum might, perhaps, be used 
to advantage. The presence of some purple 
sap in the column of Cochlioda Noetzliana 
may or may not complicate matters. It all 
depends whether the column colour is 
inherited independently of the rest of the 
flower or not ; if a form can be secured in 
which the usual violet-purple colour m the 
column is absent this possible complication 
might be avoided. 
In the making of a Scarlet Cattleya the 
elimination of the rose-purple sap of the 
Cattleya is equally important in order to get 
the true colour, and it would therefore seem 
advisable to use with the Sophronitis the 
albino forms, e.g., Cattleyj Moss.o? Wageneri 
and C. Gaskelliana alba. If the (C) albinos 
of C. Warneri and C. Schroder^E are used 
care must be taken not I0 put them in the 
same experiment in which the (R) albinos of 
C. Mossiae and C. Gaskelliana are used, other- 
wise the sap-colour will appear. 
In place of Sophronitis grandiflora, the 
species Laelia harpophylla or L. cinnabarina 
might be used, but the result would be a 
lighter and more orange shade of scarlet. 
In the making of a self yellow Cattleya, 
the species Laelia Cowanii or L. flava would 
appear to be the best to use. C. Dowiana 
aurea and L. xanthina being bicolors would 
be unsuitable for breeding- selfs. It is rather 
curious that while the yellow colour of C. 
Dowiana aurea is recessive to the rose-purple 
colour of the Cattleyas, the yellow colours of 
L. Cowanii, L. flava, and L. xanthina are 
dominant, though in most cases the dominance 
is somewhat imperfect. The Fi results sug- 
gest that the colour of the front lobe of the 
hp may be inherited independently of the 
rest of the flower. The bicolors C. Dowiana 
and L. xanthina give purple or crimson lips 
when mated with albinos, while L. Cowanii 
and L. flava apparently do not ; but L. flava 
when mated with the coloured forms of certain 
species gives a crimson or purple lip, while 
with other species it gives a self yellow flower. 
L. Cowanii so far appears to give all yellow 
selfs with both coloured and albino forms. 
In the making of a self yellow Cattleya, all 
these details will have to be carefully con- 
sidered by the breeder in choosing his 
matings. It may be interesting to note that a 
rather curious point will arise in the making 
of these Scarlet and Yellow Cattleyas. 
Technically, according to the present rules, 
the Scarlet Cattleyas bred 111 the way sug- 
gested would belong to the genus Sophro- 
Cattleya, while the Yellow Cattleyas would be 
classed as Laslio-Cattleya. But, as the writer 
pointed out some years ago, the generic 
characters which separate Cattleya from 
La?lia, i.e., four pollen-masses and eight pollen- 
masses respectively, themselves behave as 
Mendelian characters and segregate in Fi in 
the same way that specific and varietal char- 
acters do. Consequently, some of the Scarlet 
and Yellow Cattleyas raised in these experi- 
ments will have four pollen-masses only, and 
in that sense would be indistinguishable 
from the s}stematist's true Cattleya. Being 
homozygous they would also breed true to 
that character. In view of this it hardly 
