May 17, 1912] 
the New Era variety these dots are very 
numerous, while in the variety known as 
Taylor the dots are more thinly scattered, 
with considerable areas free from them. 
Certain other varieties, notably the whip- 
poorwill, are characterized by a peculiar 
mottling of the surface in which irregular 
areas of darker shade are separated by 
lighter areas, apparently of the same color- 
ing material, but less dense. There are 
also two genetically distinct types of 
“feve,’’ the data concerning the inherit- 
ance of which have already been pub- 
lished.* Finally, certain varieties are 
characterized by irregular longitudinal 
stripes on the sides of the seeds. We have 
thus, in all, 13 pairs of Mendelian char- 
acters relating to the seed coat. 
Except the purple color, the inheritance 
of all the above characters has been worked 
out quite completely. Purple was intro- 
duced into my crosses from a variety that 
had only a small color patch about the 
hilum, and was not recognized as a distinct 
color type until after the crosses had been 
made. It happened that the crosses made 
were not such as to bring out in full the 
relations of this purple color to the other 
Mendelian factors involved. Its inherit- 
ance will be investigated later. 
Without postulatmg anything whatever 
as to the nature of the Mendelian factors 
involved in the development of these colors, 
I will first set forth some of the more im- 
portant facts that have been worked out 
concerning which factors must be present 
in order that a given color may develop. 
Since we do not know what these factors 
are, and hence can not give them names 
based on their nature, and since we do 
know their behavior in inheritance and the 
relation of their presence and absence to 
the development of the colors, I shall give 
4Amer. Nat., Vol. XLV., No. 537, September, 
1911, pp. 513-24. 
SCIENCE 
763 
them provisional names based on their be- 
havior and their relations to color develop- 
ment, and then, for convenience, reduce 
these names to mere symbols. 
In no ease is there a single Mendelian 
factor that can by itself give rise to colored 
seeds. I have been able to demonstrate 
that red, buff and brown seeds each require 
at least two factors. They may require 
others, but if so these others have not been 
detected. Black requires three factors and 
blue three. 
If all these factors were distinct, these 
five colors would thus require twelve sepa- 
rate factors for their development; but the 
whole number of these factors for color 
(omitting purple) is only six. The same 
factor takes part in the development of 
more than one color. In fact, there is one 
that is common to the whole series, and 
when this factor is absent white seeds re- 
sult, even if the remaining five are all 
present. We may designate this factor as 
the general color factor, and represent it 
by the symbol C. If we represent the 
other five factors by the letters R (red), 
U (buff), Br (brown), B (blue) and 
N (noir= black), the conditions required 
for the production of each color in this 
series, as shown in the breeding experi- 
ments, is shown in the following table: 
CONDITIONS REQUIRED FOR DEVELOPMENT OF COLORS 
Factors Factors Having Factors Which Must 
Colors Necessary no Effect be Absent 
White R,U,Br,B,N C 
Red C,k NorB U, Br, B or N 
Buft C,U R,N or B Br, B or N 
Brown C, Br R,U,B N 
Black C,Br,N &,U,B 
Blue C,N,B IRS OP Br 
In this table the names of the colors are 
given in the first column. In the second 
column are given the factors that must be 
5These do not affect blue, but are visible with 
blue. 
