248 MendeVs Laivs of Alternative Inheritance in Peas 
and secondly how enormously the range of vai'iation included under the two 
differs. 
The colour of Stratagem is interesting because of its intermediate condition 
between the extremes observed in the group of races to which it belongs. As 
will be seen from Table IV., piebald seeds are rarer ihan in any of its allies, 
and the great majority of seeds fall into the colour-groups 2 and 3*, with a very 
small percentage of seeds so green as Group 1, or as yellow as Groups 4 and 5. 
The general impression produced by a mass of these seeds, when stripped of 
their coats, is certainly green ; but it is a far yellower green than that of either 
Telegraph or Pride of the Market; and I think tliere can be no question that in 
Stratagem a blend of green and yellow has been inherited, and fixed by a 
process of selection. 
(4) Telephone, in the samples given to me by Carter and Co. or by Sutton 
and Sons, is fairly uniform in shape, being practically always as wrinkled as 
Figs. 5 and 6. It appeai-s however from Tschermak's account (No. 27, p. 480) 
that samples which he bought in Ghent and in Quedlinburg were more variable, 
and contained smoother, more oblong forms. The same thing is true of a sample 
which I bought in Oxford from Messrs McGreal and Co. The colour of all the 
samples I have seen (from Vilmorin-Andrieux, Carter, Sutton, and McGreal) is 
exceedingly variable. Every one of the six colour-groups is abundantly represented, 
and indeed the scale of colours, which was found suitable for the whole series of 
races, was selected from this race alone. There can be no question that self- 
coloured seeds of every shade intermediate between the darkest green of Fig. 1 
and the most intense orange-yellow of Fig. 6 occur in every large sample 
of this race, while piebald peas, with every degree of difference between the 
colours of their patches, occur also. Groups of piebalds, illustrating their main 
characters, are shown in Figs. 13- — 18 on Plate I. The colour variation of this 
race, together with Tschermak's description of its behaviour when crossed as 
a green-seeded pea with exceptional dominance, led me to tabulate the colours 
in several samples. The results are given in Table IV. ; and while the different 
samples differ considerably among themselves, they all show the same general 
characters. In all of them the range of colour is the whole range of the scale, and 
the groups of intermediate colour occur most frequently ; in all of them there 
is a large percentage of piebald individuals. It is evident that in a very 
large percentage of these individuals the characters " green " and " yellow " are 
perfectly blended, while others present a mosaic or piebald character ; but even 
in these piebald individuals the patches of colour cannot, in some cases, be spoken 
of as pure green or pure yellow ; the pea may exhibit a blend of green and yellow 
on every part of its surface, but the proportions in which the two are mixed may 
differ in the two cotyledons, or in different parts of the same cotyledon. The 
suggestion that yellow and green are mutually exclusive alternatives, or that 
* As will be seen from Plate I., Figs. 7 — 12, many seeds of this race are intermediate in colour 
between Figs. 2 and 3. 
