314 
Breeding Experinmits with Foxgloves 
By reference to the horizontal line of totals at the bottom of the table it will 
be seen that among the purple offspring there are none under the first three units 
of the colour scale, 0-4, 5-9, 10-14. There would also, appear to be discontinuity 
at the other end of the scale, but it is very doubtful as to how far reliance can be 
placed on the colour-scale where the depth of the fluid is considerable. 
Taking the families altogether there were 139 purple plants and 41 white. 
The Mendelian expectation may be calculated for each family on making the 
following assumptions: purple colour dominant to white; C (2) pure recessive; 
A, 6'(1) and C'(3) heterozygous dominants; B{Z), B (4), A^(l) and iV (2) homo- 
zygous dominants. 
Justification for these assumptions is given by the consistency with which 
the different parents act in different pairings. The sum totals of the calculated 
expectations for all the families are 131 purple and 49 white. It will be seen that 
these figures are in very fair agreement with the experimental result which is 
139 purple and 41 white. This may be regarded as a typical Mendelian result. 
A point of interest is to ascertain as to how far the various intensities of the 
colouration among the purple plants are inherited. It is possible that some 
observers would regard the different intensities of the purple colouration in a 
family as representing mere fluctuating variations which could not be inherited. 
According to the pure-line theory the different intensities of purple colouration 
occurring within the limits of a family resulting from self-fertilization cannot be 
inherited. It will be possible to test this theory in the generation of plants 
now being raised, but the present material cannot give a definite reply to the 
question. 
An inspection of the above table (p. 313) will show that as we pass to the higher 
mid-parental colours there is a gradual rise in the colouration of the families. 
This implies that the various shades of purple are not distinct types, but that 
they can blend perfectly with one another. 
The fact that in closely similar matings the famihes raised show a regular 
progressive rise in the intensity of colouration with the rise in the mid-parental 
colouration does not favour the expectation that any support for the pure-line 
theory will be forthcoming from the present investigation. 
It is desired to find the correlation with respect to the intensity of colour 
between the purple parents and the purple offspring. For this purpose, in order to 
have strictly homogeneous material, the parents should be homozygous dominants. 
The material will not permit such a calculation. The purple plants of the different 
families include both homozygous dominants and heterozygous dominants. Now 
the homozygous dominants are found to be darker in colour than the heterozygous 
dominants, and so in any family consisting of an equal number of heterozygous 
