70 
In this case, the 16,384 possible reproductive cells (pollen 
grains or ovules) of hybrid X may be assumed to differ from each 
of the 16,384 possible reproductive cells in hybrid Y. This would 
make the case more complicated, for then in the cross between 
X and Y the possible combinations would be for the first chromo- 
some aa, aA, AA, al, Al, al, AI, 11, 11, II, or a total of ten, the same 
number of combinations for the second, and so forth, making the 
possible number of different kinds of seedlings resulting from a hy- 
brid made up from four such species 10** or 100,000,000,000,000. 
Whether we would ever find four gladioli as radically different as 
is here assumed, and still find it possible to intercross them is not 
sure. So this is a purely hypothetical case—but even so the calcu- 
lated results seem rather astounding—and just show what sur- 
prising possibilities there are along these lines. This computation 
takes account only of the different seedlings that might arise with 
a normal segregation of chromosomes. If chromosomes split ab- 
normally, causing “crossing over” of characters normally linked 
on one chromosome then the numbers of possible combinations 
might be indefinitely increased. Of course, in genetical studies of 
gladioli or other plants nobody has attempted thus far to follow 
through as many as fourteen independent heritable characters 1m 
the manner indicated ; no one garden or force of scientific workers 
would be large enough to make it practicable. At most, breeders 
work with two, three, or four characters of their plants, and even 
so they must grow thousands of seedlings before they can estab- 
lish the manner of inheritance of these few traits. 
The possibilities of different kinds of gladioli among seedlings, 
though, are so numerous that the tens of thousands of different 
seedlings raised each year by breeders are only a small fraction of 
the variations that may be expected among the modern hybri 
strains, descended, not from four different species, but from at 
least ten. Nor is that the end ; there are fully one hundred mor? 
gladiolus species that certainly have not yet been used in breeding: 
There may not be as many as 100,000,000,000,000 possibilities 
from any one set of hybrids, involving four species, because the 
Species may be more nearly alike than was assumed in this calcu- 
lation. But with so many different species that will readily inter” 
cross, the possibilities in hybrids are far greater than have ye 
been attained. 
New York BOTANICAL GARDEN 
