GENETICS: CLAUSEN AND GOODSPEED 
241 
another. In Drosophila, for example, the development of the normal 
abdomen under certain environmental conditions in spite of the pres- 
ence of the factor for abnormal abdomen in the reaction system indi- 
cates the existence of compensatory relations among the factors of the 
system. Such compensatory relations are even more strikingly evi- 
denced in the case of maize seedhngs of the yellow-green chlorophyll 
reduction type. Normally these die, but under favorable conditions 
the system is able to overcome the disturbance incident upon the pres- 
ence of the chlorophyll reduction factor and to go on and develop the 
normal chlorophyll coloration in the plant. Similarly, the lethal effect 
of changes in certain loci, the similarity in effect of different changes in 
the same locus displayed in multiple allelomorphism, the apparently 
universal significance of the multiple-factor conception of character 
development, and a variety of other considerations indicate that impor- 
tant p>hysiological relations exist among the loci of the system, and that 
character expressions depend upon the reaction-system relations of 
Mendelian factors. The product of somatogenesis, the individual, repre- 
sents the reaction end-product of such a physico-chemical system work- 
ing under particular conditions; the specific hereditary differences 
between individuals of the same species indicate particular differences 
in some one or more elements in such a reaction-system. Normal 
Mendelian behavior, then, would follow as a result of hybridization 
phenomena involving a contrast between a relatively few particular 
differences within a reaction system which is fundamentally identical 
in the races under consideration. If in contrast to this type of behavior 
it should be possible to secure contrasts of fundamentally different 
reaction systems, then conceivably the elements, although playing defi- 
nite parts in their own systems, might fail to estabhsh the harmonious 
inter-relations which are necessary for normal development and repro- 
duction. Such incompatibility of elements would give rise to a peculiar 
type of behavior in inheritance which could not well be accounted for 
by the customary formal treatment based on the Mendelian viewpoint. 
The experimental data which we have collected seem to indicate that 
such a situation actually does obtain in certain cases of hybridization 
between distinct species. 
For ten years a number of species and varieties of Nicotiana have 
been grown in the University of California Botanical Garden. Among 
many others this collection has included N, syhestris and a considerable 
array of varieties of N. Tabacum. The varieties of Tabacum display 
notable morphological differences throughout — differences so marked 
that to regard them as distinct species would be entirely justifiable, 
