180 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
SUMMARY. 
Bringing together the essential facts for the five colonies analyzed in detail 
(table 1143), the two series of Papenoo and Paraura support the conclusion that the 
red color is a Mendelian dominant in its relations to yellow; the series of Papeiha 
and Ahonu are not conclusive either way, while the results only in the case of the 
Oopu series distinctly favor the opposite contention. 
Tasie 1143.—Partula otaheitana rubescens. Summary of the essential results 
of analyzing the heredity of the red and yellow colors in Mendelian terms. 
Red color assumed to Yellow color assumed to 
be dominant. be dominant. 
Valley. 
Proporioneyor Percentage of Propertioneyor Percentage of 
DD:DR:RR error in inde- DD:DR:RR error in inde- 
pendent test. ; pendent test. 
Oxo bocc00c 13:34:27 14.4 12388 WU 8.1 
Papeihay. 4... 738 NS) 8 SK0) 19.6 6:12:10 18 
Raralramerenrer BS O83 38 0 Gg fe 4! U 3@) 
Papenoo...... 5:36:18 0.7 impossibility 
INDOYMR o6.00006 83g iil s ily 22 G9 iil 9 ital 19.3 
As a rule, the colonies are relatively stable in their color-composition, which 
would scarcely be the case if the yellow and red were not hereditarily determined. 
Of course small changes would be expected from year to year, unless the relative 
numbers of DD, DR, and RR individuals had reached such proportionate relations 
as to be repeated identically in generation after generation; but the natural fluctua- 
tions would be small in an equilibrated population of long standing. 
The inheritance of the ground-color is alternative, so far as the observations 
may be trusted. Among the embryonic snails, the colors are attenuated, and 
errors in determination have been made without doubt, especially in the case of the 
Papeiha embryos. But every single original observation has been made by one 
individual observer in obtaining the basic facts, with no bias other than the 
endeavor to assign a shell to the yellow or to the red class. Not until the tables 
were completed did it appear that the two classes of embryonic snails were pro- 
portionately represented as in the adult populations or otherwise. 
In devising the method for discovering a Mendelian behavior of the color- 
factors, almost insuperable obstacles were encountered, notably the small numbers 
of young produced by a single parent, the impossibility of ascertaining the gametic 
composition of the unknown mate of a gravid adult, and lack of knowledge as 
to the relative numbers of DD, DR, and RR snails, although in the last connec- 
tion it was evident that one class would be RR and the other would be DD+DR. 
According to the procedure adopted, it was first assumed that the yellow class repre- 
sented RR, and the relative numbers of the three kinds of adults were then ascer- 
tained and were tested by the correspondence between the observed and expected 
relative numbers in the other color-class, by an entirely independent analysis. Sub- 
sequently the converse assumption was made, and the results were compared with 
the first conclusions as to their consistency. 
