()2 
THE ORCHID WORLD. 
another point which may be of some interest. 
Records of the flowering of seedHngs from 
selfed hybrids are exceedingly rare in 
Orchids. At one time the writer thought that 
this might be due, on the one hand, to the prac- 
tical prejudice against in-breeding, or, on the 
other hand, to the idea that nothing new 
might be expected from selfing. Now, how- 
ever, after attempting to self several hundreds 
of different flowers of Cypripedium and other 
h\-brids during the past ten years, the writer 
has come to the conclusion that self-sterility 
m hybrid Orchids has to be reckoned with, 
being apparently the rule rather than the 
exception. For instance, since publishing my 
Mendelian experiments with Cypripedium 
Hera m 1903, I have from time to time 
endeavoured to self the individual Fa hybrids 
as they flowered, but in every case but one, 
when the pods formed, they afterwards proved 
to be empty. A few seeds, however, were 
secured from one pod of selfed C. Hera 
punctatissimum and a single Fj plant has 
flowered which bred true to its spotted parent 
in so far as the spotted character was 
concerned. 
On the other hand, Mr. R. A. Rolfe has 
obtained a number of seedlings at Kew of 
selfed Epidendrum kewense, several of which 
have flowered, and other odd records are to 
be found during the past twenty years where 
selfed Orchid hybrids have produced plants 
which have flowered. These cases, however, 
would appear to be exceptional. 
In i8g8 the writer showed that sterility in 
hybrid Cypripediums was due in certain cases 
to a loss of power in the pollen of hybrids. 
For instance, species x species produced 95 
per cent, of fertile pods, while species x 
hybrids gave only 56 per cent. ; again, hybrids 
X species produced 89 per cent., while hybrids 
X hybrids gave only 59 per cent. It is pos- 
sible that the same cause may be concerned 
in the self sterility of Orchid hybrids 
generally, plus some other factor unknown. 
Whatever difficulties may be met with in 
selfing hybrid Orchids, no such difficulty is 
apparently encountered in selfing species and 
varieties. 1 have found no difficulty in raising 
seedlings of selfed species and varieties of 
I December, ujiz. 
Cypripedium, and many records are to be 
found m this and in other genera in Orchid 
atera iure. In view, however, of the great 
successes achieved by Orchid breeders in the 
hybridisation of diverse species and genera, 
it is hardly likely that selfing will be resorted 
to now to any great extent, either in species 
or hybrids. 
As we have seen, cross-breeding is a far 
more effective method of analysis than se'fing, 
even from the point of view of genetics. The 
crux of the whole ])roblem, however, lies m 
the identification of individual plants Unless 
this IS done all the rest is vain. Unfortu- 
nately, with a few brilliant exceptions, our 
multitudinous Orchid records and labels are 
almost useless for this purpose. Seme of 
these brilliant exceptions consLSt of certain 
unique hybrids to which their owners gave a 
distinctive name, proudly refusing to conform 
to the technical rules laid down by the 
authorities in nomenclature (and by the irony 
of fate the writer happened to be one of the 
latter). Many of these names can be traced 
in the Orchid Stud Book as synonyms, and it 
is possible that the natural conservatism of 
Orchid breeders has caused them to retain 
many of these old names on their labels. If 
such is the case, the identification of some of 
our best stud individuals ma\' not be so 
difficult after all. 
In most cases, however, the adoption of 
genetic methods will necessitate the making 
over again of the best hybrids, using as 
I)arents the best individuals of the best 
varieties of the best species of the best 
genera. The primary hybrids thus obtained 
would be pedigree hybrids, and the best indi- 
viduals of these might be further utilised in 
accordance with the particular aims that the 
breeder may have in view. In this way the 
Orchid breeder would be able to build up a 
pedigree strain made up of .superior indi- 
viduals, and the reversions and recombina- 
tions that he would obtain among the 
secondary hybrids would also tend to be 
superior to those bred in the ordinary wa}'. 
In breeding on these lines the Orchid 
breeder will soon discover for himself that 
while a few stud individuals are homozygous, 
