228 
Allen. — Nuclear Division in the 
granules may themselves well possess an inconceivably complex structure ; 
and it is not yet practicable to attempt to identify the lowest observable 
order of elements in the spirem-thread as the bearers of the apparently 
simple unit qualities manifested by the organism. 
If the parents, the transmission of whose qualities to their offspring 
is to be studied, are of the same race, and so differ from each other only 
in the minor respects which fall within the range of individual variation, 
the study of heredity offers serious difficulties. It is, therefore, when the 
parents differ from each other by characters of a racial or of a higher order 
that the inheritance or non-inheritance of special parental qualities is most 
easily determined ; and consequently, the experimental facts which are 
of most value for correlation with, and elucidation of, the results of 
cytological study are, in general, those derived from a study of cross- 
breeding between individuals of different races, varieties, species, or even 
genera. 
It has been shown that the parental hereditary substances (idioplasms) 
to all appearances remain separate during the life of a sexually-produced 
individual, so that each cell of such an individual contains the physical 
basis for two complete sets of individual qualities. As to any particular 
quality, then, with respect to which the parents differed — e. g. shape of 
leaves — there are in the offspring two different, and therefore more or less 
conflicting, hereditary tendencies. As a consequence of the joint action 
of these two tendencies, some resultant quality must appear in the offspring, 
whose relation to the corresponding qualities of the parents may be shown 
in one of two ways — either the resultant quality may be intermediate 
between the parental qualities ; or one parental quality may entirely 
predominate over its opponent, and the offspring in this respect exactly 
resemble one parent. Observation shows that both of these hypothetical 
cases are actually realized. 
A well-known instance of a blending of parental characters is furnished 
by the hybrid Drosera obovata , whose leaves are intermediate in shape 
between those of the parents, D. rotundifolia and D. longifolia. Mendel (’65) 
found that in various Pisum hybrids the time of flowering was almost 
exactly intermediate between the times of the parents. Macfarlane (’90) 
observed in numerous cases that the blending extends to matters of cellular 
structure ; for example, in a hybrid Hedychium the starch-grains were 
intermediate as to size and shape between those of the parents. A different 
sort of blending, apparently clearly due to the independent existence 
within the cell of the parental idioplasms, is that noted by Hildebrand (’89) 
in a hybrid between two species of Oxalis ; in this instance both forms of 
hairs characteristic of the parent species arise from a single epidermal cell. 
What may perhaps be considered a special case of the blending of 
parental characters is that in which the resultant quality appears to be, 
