IN THE TROPICS. 
325 
of characters concerned. On the other hand, two parents 
showing the recessive character can produce nothing but 
recessive offspring in the simple Mendelian case. 
Pearson (54) has worked out more completely a “ genera¬ 
lized theory of the pure gamete” in cases where dominance 
is wanting, and an indefinite number of pairs of allelomorphs 
are concerned." On this theory Pearson finds a value of 
one-third for parental correlation (in the case of bi-sexual 
reproduction without preferential mating), a value identical 
with that obtained by Galton in his original investigations 
upon the inheritance of stature (37). 
Pearson finds that the statements : firstly, that the average 
value of parental correlation deduced from recent biometric 
observations upon all kinds of material appears to be from *45 
to *5 ; secondly, that this correlation appears to vary slightly 
from character to character ; and thirdly, from species to 
species, are inconsistent with the theory of the pure gamete ; 
which throws the “ Mendelian ” back into the position of the 
biometrician of 1885.| 
We have seen, however, that in certain cases the simple 
form of Mendel’s theory has been proved to hold good, a fact 
of which Pearson takes no notice. On the other hand, not 
even a “ Mendelian ” supposed, even in 1900, that Mendel’s 
Law held good for all characters in all species. But Pearson 
ignores the whole mass of non-biometric work of the last 
four years with its examples of irregular dominance and 
latency of characters. These factors, as well as the effect of 
conditions upon posLgerminal development, are entirely 
indeterminate in the cases for which biometrical determina¬ 
tions of parental correlation have been made. It seems 
* Presumably in determining a single measurable external character¬ 
istic, since otherwise the results would not be comparable with those for 
stature. 
f Pearson here seems to overlook the fact that he himself, no longer ago 
than 1900, found theoretical reasons for using the value *3 to represent the 
intensity of parental correlation ; a figure still further removed from the 
“ cluster point of existing measurements ” than the value *. (See Grammar 
of Science, Ed. II., p. 459). 
8(10)04 (25) 
