96 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
without many qualifications, chiefly because the several color-classes are not uni- 
formly represented in the same proportionate numbers in all associations, where- 
fore the opportunities for mutual intercrossing are anything but the same in all 
parts of the island. Although an analysis on the basis of individual associations is 
necessary, yet the collective statistics show some correspondence with the facts 
relating to the adolescents, in so far as the majority of the young of phea and 
marginata belong to the darkest group (type c) while those of unicolor and bicolor fall 
mainly in the lightest group. The young of mitella are more frequently whitish and 
uncolored than would be expected, while the embryonic shells of mitella-rubra are 
dark in an unlooked-for number of instances. 
TABLE 48.—Partula gibba, Guam. Comprehensive statistics of heredity of color. 

Young. 
Parent class. " Type B, Type ¢, 
ype a, . 
Nehltteli: yellowish, brown- 
tinged apex. corneous. 
WMICDION 0 6 yaaa e nee eee ase earee 139 Se 
bicnlar ies gk ees eee Oe ox wes cee Lak 53 ee 
mrtelle iv t.ka'ss ic cimis ince wei s 947 333 2 
mnitella-rabrad. ssn. .hepw cobs Gah 77 97 78 
ee ee ae ee ee 95 23 127 
CUSTENOR Se oo chee oe ha ee a ewe 34 13 pai Ee 
VEEDBIR wisc5 crea elec ean la ee eoleioa a 183 2 1 
INATPINAER vaio n. a oie mae ernie eae Res a 7 
QUAL ars crx cptnircic ata tee 1,646 471 215 



The detailed analysis begins with the associations which comprise only mitella. 
Combining the material from the upper Ylig, Cemetery, and the two Presidio local- 
ities, the number of young of type a is 193, while the young of type b amount to 70. 
It is obvious that the definitive mitella mode of coloration is very slow to develop, 
for certainly we would not be justified in concluding that the 193 young with the 
type a colors would become anything else but mitella adults. 
Passing to Tarague, we have an association that comprises 5 color-classes, 
whose young are distributed among the 3 groups as shown in table 49. The sig- 
nificant points are that all of the embryonic individuals from unicolor and bicolor 
adults belong to type a, that very few young of mitella display tinged apices, and 
that many of the young of mitella-rubra and phea are dark-colored. The whole 
series of adults taken at Tarague amounted to 292 with the proportionate repre- 
sentation of the color-classes as recorded in the census table (table 29), while the 
replacement generation amounts to 288, which is very nearly equal numerically. 
Assuming that the association is substantially stable, the embryonic young will 
develop into adults of the various kinds with about the same numerical relations as 
those of the parent generation. Now, the adults of the lightest classes numbered 162, 
while the type a young amount to 259; clearly, therefore, some of these embryonic 
individuals are destined to become other than unicolor or bicolor adults and, from 
the considerations noted in the foregoing paragraphs, the presumption is that most 
of such young would have become mitella. ‘Thus intercrossing of the light adults 
