308 Saunders. 
explain the facts regarding singleness and doubleness). But GoLD- 
SCHMIDT’s scheme of sex-limited inheritance breaks down here also, 
As in the case of doubleness it failed to provide an explanation of 
the excess of doubles obtained from types which are eversporting in 
regard to doubleness, so in the case of the one type which is ever- 
sporting in regard to plastid colour, viz. the sulphur-white, it fails 
to meet the facts, for it does not account for the occurrence of the 
small percentage of whites observed to occur among the doubles. In 
order to meet this difficulty GOLDSCHMIDT has recourse to the 
explanation of “crossing-over”. But to appeal to crossing-over is 
equivalent to an admission that the result in question is not explicable 
by his theory as it stands. It would be possible to dispose af almost. 
any inconvenient fact which is not in accordance with expectation in 
this convenient way, for, supposing it to occur, we are without any 
knowledge of the way in which the process would be regulated. 
After referring to MoRGAN’s view that certain unexpected results 
obtained by him with Drosophila, which seem otherwise inexplicable, 
must be attributed to the occurrence of crossing-over, i. e. to the 
exchange of substance between the two members of a chromosome pair, 
GOLDSCHMIDT concludes his argument with the astonishing statement 
that we therefore have the right (the italies are mine) to assume 
that in the present case (i. e. Stocks) W and w are occasionally inter- 
changed, that crossing-over occurs, and therefore unexpected com- 
binations become possible”. But on my view there is no need to regard 
the production of a small percentage of double whites by the sulphur-- 
white strains as an “unexpected combination”, which needs to be 
accounted for by the somewhat indefinite and unaccountable process. 
of “crossing over”. Their occurrence can be satisfactorily explained on 
the supposition of partial coupling of the kind already described (see 
above, and Journ. of Genetics, p. 323), an explanation which has been 
shown to be adequate in the case also of doubleness(t), not to mention 
the number of other cases, notably, e. g. that of the Sweet Pea(?), in 
(1) As lending some support to his suggestion of crossing-over GOLDSCHMIDT 
quotes the results obtained in one of my experiments with a certain individual 
plant K. The case of this individual is fully discussed in my original paper (see 
p- 315 and Table II), and I need not repeat the whole argument here. The point of 
importance is that the evidence clearly indicated that this supposed eversporting 
cream was of cross-bred origin. Since we are ignorant of the exact constitution of 
this individual we cannot safely use the results obtained from it or its descendants. 
as a basis for argument in the present discussion. 
(?) See note p. 299. 
