238 Mendel's Laws of Alternative hiheritance in Peas 
Mendel's work, and the result of crossing Peas has lately been described by Correns 
(Nos. 3 and 4) and in greater detail by Tschermak (Nos. 27 and 28). Correns 
confirms Mendel's statements concerning the dominance of yellow cotyledons, but 
Tschermak makes a more detailed statement, which does not so fully agree with 
Mendel. It is not quite easy to follow Tschermak's account, because he does not 
describe all his very numerous and careful experiments in siich a way that one can 
be sure how many hybrid peas he observed. He certainly crossed between 80 and 
100 flowers, belonging to green or yellow-seeded races, with pollen from plants with 
seed of the opposite colour, obtaining between 300 and 400 hybrid seeds. Of 
these hybrid seeds about 40 were not distinctly yellow, so that 90 per cent, of the 
hybrids exhibited the dominant character. Of the seeds which did not exhibit 
this character, some were yellow with green patches, eight were green, four were 
"yellowish green," and five are described as showing " Griin und Uebergange von 
Gelb zu Grlin," so that while some 10 per cent, of the hybrids did not exhibit 
dominance of yellow, some 2 or 3 per cent, exhibited a close approximation to the 
character of the " recessive " parent. A further case is recorded by Tschermak, 
where a plant of the yellow-seeded variety Buchsbaum, growing in the open, 
produced a pod in which every pea except one was green, the exception showing a 
little yellow. One of the green seeds was sown, and the plant produced fifteen 
yellow and three green seeds, thus behaving like a hybrid. Tschermak considers 
that this is a case of accidental crossing, with dominance of green, and although 
some of the green peas, produced in his experiments, may have been due to 
accidental self-fertilisation, he regards some of them, at least, as hybrids (cf. 
No. 28, pp. 663 — 664). It should be said that some of the crosses referred to 
were made between the (yellow-seeded) P. arvense, var, Graue Riesen, and green- 
seeded varieties of P. sativum. In all such crosses the law of dominance of 
yellow held absolutely. 
These results show clearly enough that the law of dominance is, as Tschermak 
says, not absolutely true of cotyledon colour, and as will presently be shown, the 
exceptions to the law, which he observed, form a very large percentage of the 
total result obtained when certain races were crossed. 
The case of cotyledon colour has been considered first, because the evidence 
with regard to it is more favourable to Mendel than is the evidence touching other 
seed characters. The shape of seeds, whether smooth and rounded or irregular 
and wrinkled, is even more difficult to express in words than the colour of their 
cotyledons, and the varieties appear to be even less constant. The evidence 
against the universal validity of the law of dominance is here much stronger than 
in the case of cotyledon-colour. Two striking cases were observed by Rimpau 
(No. 26, pp. 36, 37), who crossed the smooth-seeded race Victoria with two wrinkled 
races, Knight's Marrow and Telephone. He made each cross in both the possible 
ways, and found the second hybrid generation dimorphic, as usual. From the cross 
Victoria % x Knight's Marrow ^ he obtained round and wrinkled seeds of the 
second generation, as Mendel's statements would lead us to expect. The wrinkled 
