148 
Variation and Inheritance in Aphis 
We know from Driesch's experiments that the mere position of a cell in the 
blastosphere determines its fate, and that the blastomeres are at first, so to speak, 
merely plastic masses which are welded into an organism by the life of the 
structure as a whole. The failure of inheritance must be regarded as of the same 
nature as the failure in the growing organism to produce the normal relative pro- 
portions of the different parts of the body. 
And finally, with regard to the special subject of this paper, inheritance in 
parthenogenesis. From the nature of the case, since there is only a female 
parent, the inheritance cannot be a blend of maternal and paternal characters. 
If it were " exclusive inheritance " without reversion the coefficients ought to be 
far higher than in sexual forms. But this is not the case. The inheritance could 
be either of the nature of a blend of ancestral characters, or the "exclusive" 
transmission of ancestral characters, that is reversion. 
In the accompanying table in the second column I give the theoretical values 
deduced by Professor Pearson* for exclusive inheritance, without reversion, in 
the third column the values for blended inheritance or exclusive inheritance with 
reversion and 7 = 2"35f. In the fourth and fifth columns are the values found for 
eye-colour in man and coat-colour in the horse, and in the sixth column the 
TABLE XXIII. 
Kelationship 
Exclusive 
Inheritance, 
no reversion 
Blended Inherit- 
ance or Exclu- 
sive Inheritance 
with reversion. 
7 = 2-35 
Man. 
Eye-colour 
Horse. 
Coat-colour 
Aphis and 
Daphnia. Mean 
coefficients for 
all dimensions 
Aphis and 
Daphnia. Mean 
coefficients for 
ratios only 
Parental 
■50 
•40 
•49 
•52 
•41 
•45 
Grandparental 
•25 
•20 
■32 
•33 
•24 
•25 
Fraternal 
•4 to 1-0 
•66 
•47 
•63 
•66 
•64 
mean coefficients for aphis and daphnia. In the case of the parthenogenetic 
animals the parental correlation compared with the results from the sexual forms 
is low, and the fraternal correlation distinctly high. If we exclude the less 
reliable absolute dimensions and only consider the ratios, the parental correlation 
is raised (seventh column). The question to be decided by further investi- 
gation is whether these differences are real or are merely the effects of the 
I'eaction of the organism to its environment. So far as the present investi- 
gation goes there is no increase, rather a decrease is shown in parental in- 
heritance as we pass from sexual to parthenogenetic forms. There is possibly an 
increase, but we cannot as yet assert it definitely, in fraternal correlation. On the 
whole the numbers for both sexual and parthenogenetic inheritance lie broadly 
within the same limits. 
* Phil. Trans. Roy. Soc. Vol. 195 (a), p. 119. 
f Proc. Roy. Soc. Vol. 66, 1900, p. 149. 
