
Shell. Aperture. 
Series. No. 
Length. Width. Proportions. Length. Width. Proportions. 
Saucio mm. mm, p. ct. mm. mm. p. ct. 
Havenicc. cdo ee 37 0.6637 +.0520} 0.3366+.0263) 1.8576+.1456| 0.3582+.0281) 0.3001+.0235| 1.9322+.1514 
pallidat.: £260.00 ne 18 .6557 +.0737 .3047 +.0342| 2.1315+4.2395 -4123 +.0463 .3207 +.0360| 1.3933 +.1566 
bicestatal 2. 350-24 boke 2 .1500 +.0505 -1000 +.0337| 0.5000 +.1686 0 -1000 +.0337| 1.5000+.5058 
All tonto eee 57 .7004 +.0442 .3290+.0208| 2.0424+.1290 .3730+.0235 .3114+.0196) 1.8203+.1150 
Aniguac: 
flavea 21008 Gur hE sais 26 .6068 +.0567 -3595+.0370} 1.9154+.1791 .3730 +.0348 .3099+.0290) 2.2182+.2074 
pallidatWie hie elses 11 ~ .5246+.0754 .3699 +.0532} 1.8719+.2692 4748 +.0682 .2840+.0408| 1.7249+.2480 
AL ao eee te ee ake 37 .5904 +.0462 .4231+.0331| 1.9813+.1553 .4274+.0335 .3239+.0254| 2.1229+.1664 
Presidio, outer: 
flavea 4.52 5,.8528 es 1% 17 -5989 +.0693 .3213+4.0371| 1.9133 +.2213 -3105 +.0359 .2651+.0306| 1.4741+.1705 
pallida {Gan eeee ees 5 .5161+.1101 .3200 +.0682| 3.0066+.6413 -2039 +.0435 .1600+.0341| 0.9798+.2089 
All & SAGE eo eae 22 .5813 +.0591 .3213+.0326| 2.2100+.2247 .2949 + .0330 .2461+.0242| 1.3817+.1405 
48 




TABLE 17.—Partula radiolata, Guam—Continued. 


STANDARD DEVIATION—Continued. 









VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 

Length aper- 
ture + length 
shell, propor- 
tions. 
p. ct. 
1.5283 +.1198 
1.7786 +.1999 
0 
1.7062 +.1077 
1.3744+.1285 
1.2792 +.1839 
1.3876 +.1087 
1.3389 +.1548 
1.7435 +.3718 
1.4438 +.1468 
in widely separated localities like the Apra Region, Agafia, and Saucio, and of 
strigata-helix in equally disconnected places without intermediate representatives, 
even where the collections are large. The significance of such facts is obvious, and 
it is not necessary to repeat the arguments with regard to the relative values of 
hereditary and environmental circumstances; mutatis mutandts, what has been said 
concerning the variations in structural characters holds true for the qualities of 
coloration with equal if not with greater cogency. 
STATISTICAL CHARACTERS OF THE COLOR-CLASSES. 
The ultimate analysis of the radiolata material involves the comparative 
treatment of the subordinate color-groups within the several associations, on the 
basis of the quantitative characters of the shells. In effect, therefore, the subjects 
of the two preceding sections are combined and carried to their final stage of dis- 
cussion. The fundamental purpose of this task is to record the facts as observed 
in order that they may be available for future studies of a similar nature that may 
some time be made, and whose results would disclose the nature and extent of such 
colonial changes as would have taken place in the interim. This object is accom- 
plished by the presentation of the full data of tables 17 to 20, which comprise the 
statistics for the colonies of sufficient size to warrant their detailed description. 
While it is unnecessary to develop the argument at length, it is obligatory to 
point out that here again are abundant evidences showing that there are no causal 
effects of the environment. In the first place, under the identical conditions of one 
and the same locality, two color-classes may be distinctly different in structural 
qualities; for example, in the case of the lower Ylig series (table 18) the strigata 
shells are collectively short and stout as compared with their lighter-colored 
associates. Conversely the members of a particular color-type existing in two 
