288 Non-Isolable Races. 
ing easily produced by green plants and green branches 
by variegated ones. 
Now let us consider the yellow seedlings of variegated 
plants. They appear, it is true, to be mutants, but, as a 
matter of fact, they are the extreme variants which, how- 
ever, do not attain to their goal but perish in the attempt, 
for they are too poor in chlorophyll and are thereby des- 
tined to die early. Nearly all of them die without so 
much as having unfolded their first leaves, or sometimes 
even their cotyledons. They constitute the extreme limit 
of a long series of variegated forms, but have, so to 
speak, followed a wrong direction. They are by no 
means rare ; for instance they are well known in the 
holly, Ilex Aquifolimn, and they often result in a very 
considerable loss amongst the seedlings raised from the 
seed of variegated plants. 
It is not, however, variegated plants only which pro- 
duce such seedlings ; green plants do so only too often, 
and this even occurs in families cultivated for experi- 
mental purposes when the cultures are pure green and 
have been so for many years or did not produce more 
than an occasional variegated leaf or twig. If in such 
cases the seeds of the single seed-parents are sown sepa- 
rately the proportions in which variegated seedlings oc- 
cur in the various groups are found to vary greatly. 
Some species appear never to produce them, for in- 
stance the tricotylous races of Cannabis sativa, Mcrcu- 
rialis anmia, and Phacclia tanacctifolia which I have 
cultivated, although I have sown the seeds of several 
hundreds of individual plants separately in the course of 
some years. In other species they are very rare ; in some, 
however, the percentage of yellow seedlings is so con- 
siderable as to become a real nuisance. Thus, for in- 
