764 
present in order that a given color shall 
appear. Thus none of these six factors is 
necessary to the production of white seeds. 
For red seeds there must be present the 
factors C and R; and so on. It will be 
noticed that the factor C is necessary in all 
the colored types. 
The third column shows the factors 
whose presence or absence is immaterial in 
each color type. Thus white seeds may or 
may not possess any or all the factors ex- 
cept C. Judging from results secured by 
others in other species, there could also be 
white seeds possessing the factor C, pro- 
vided all the other factors concerned in 
color production were absent, but thus far 
I have not found white seeds of this char- 
acter in cowpeas. 
In the ease of red and buff seeded varie- 
ties it will be noticed that N or B may be 
present, but they may not both be present. 
The reason for this is that C, N, and B are 
the factors for blue; hence if both. N and 
B were present in red or buff varieties 
these would be converted into blue. Oppo- 
site blue, in the third column, & and U are 
given as factors that may be present with- 
out modifying the blue color; while this is 
true, it-is also true that these colors (red 
_and buff) can be seen along with the blue 
in blue seeds. This seems to be due to the 
fact that blue is a sap color, and is usually 
not very intense, while red and buff are 
pigments in granular form. It is inter- 
esting to note at this point that these sur- 
‘mises of mine as to the nature of these 
colors have been confirmed by Dr. Mann, 
to whose work with these pigments I shall 
refer later. 
The last column shows factors that must 
be absent in the various color types. Thus, 
if C is present, white seeds do not occur, 
unless indeed all the other factors are ab- 
sent as well. The factor U can not be 
present in red seeds, for it would convert 
SCIENCE 
[N.S. Vou. XXXV. No. 907 
them into buff. In general, omitting blue, 
each of the colors named in column one 
completely conceals, or possibly prevents 
the development of, all those above it. 
Hence the factor Br can not be present in 
red and buff seeds. As already stated, the 
factors B and N can not both be present in 
red seeds, and the same is true of buff, 
though either of them may be present if the 
other is absent. The factor N must be 
absent in brown seeds, for otherwise we 
should have all the factors for black, and 
black renders brown invisible. Similarly, 
the factor Br may not be present in blue 
seeds, since it would change their color to 
black. 
All the factors mentioned in this table 
are fully demonstrated in the experimental 
work, the full data of which will be pub- 
lished later. Not only that, but these fac- 
tors, whatever they are, can be shuffled 
about and combined in any manner desired, 
practically at will. There are no inter- 
mediate stages to be dealt with. A factor 
is either present or not present, and that is 
the end of it. The only difficulty that 
presents itself in attempting to produce 
any color desired lies in the indifferent 
factors shown in the third column of 
the table. In some eases it is necessary to 
test a variety by appropriate cross-breed- 
ing in order to ascertain what characters 
are concealed in it, before we know what to 
expect when it is crossed with certain 
others. This will be necessary, for in- 
stance, in order to ascertain whether a 
brown variety possessed the factors & and 
U. It would not be necessary, however, to 
test any variety having colored seeds in 
order to learn whether it contains the fac- 
tor NV, for if N is present the green parts 
of the plant will exhibit anthocyan, which 
they will not do if N is absent. To learn 
whether a white variety contains N we 
should eross it with brown, in which case 
