PARTULA OTAHEITANA. 179 
The difference between the expected and the observed figures amounts to 3.5 
in 44, or 7.9 per cent. Hence the first assumption as to the dominance of red is 
supported, in spite of the small numbers available for the analysis. 
ParpENoo VALLEY. 
The classified parents fall into the following groups: 
Red adults: red young, 54; red and yellow young, 23; yellow young, 53 =130 
Yellow adults: red young, 10; red and yellow young, 10; yellow young, 39=59 
When the red color is assumed to be dominant, the RR X RR combination would 
come about in 31 per cent of the 59 cases, or 18. Effecting the transfer of 3 from 
the first and third groups to the middle group, the numbers for the three genetic 
classes become 5 DD:36 DR:18 RR=5s9. As the proportions of DD:DR in the 
class of red parents are 5:36, the absolute numbers prove to be 16 DD:114 DR. 
The former will all bear red young, whatever their mates, while 1o of the latter, 
having DD mates, will belong to the same category. According to theory, there- 
fore, there should be 26 parents with red young only, and ro4 with both kinds of 
offspring. The corrected empirical results give 25 and 105, which differ from the 
expected figures by only 1 out of 130 cases, or 0.7 per cent. 
On the contrary assumption, the chances for recessive red parents having mated 
with similar RR adults are 69 per cent of the 130 cases, or 89.7—a figure which greatly 
exceeds the actual number of 54 that bear red young only, among which are some 
that must be capable of bearing both yellow and red offspring as the result of a 
mating with DR yellow individuals; thus the analysis breaks down. 
The facts in the case of the Papenoo colony, when this is taken by itself, afford 
strong support for the contention that the red and yellow colors constitute a Men- 
delian pair, of which the former is dominant. 
AuHoNnu VALLEY. 
As the collection from this valley is small, there are relatively small numbers 
available for consideration. Here the yellow-shelled snails amount to 54 per cent 
of the population. The parents with distinguishable young are as follows: 
Red adults: red young, 19; red and yellow young, 1; yellow young, 5 =25 
Yellow adults: red young, 4; red and yellow young, 1; yellow young, 26=31 
We need not enumerate the steps in the analysis. When red is assumed to be 
dominant, the proportion of DD, DR, and RR come out as 3:11:17, respectively. 
The test in the case of the red adults reveals a difference between expectation and 
observation amounting to 5.5 out of 25 individuals, or 22 per cent. When the 
Opposite assumption is made, namely, that yellow is dominant, the proportions 
of DD, DR, and RR prove to be 3:11:11, and the independent test of these relations 
in the case of the yellow adults results in a discrepancy of 19.3 per cent. The 
second assumption is favored by the lower indicated error, but only to a very slight 
degree. 
