92 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
that the rate of reproduction, equilibrated in time with reference to eliminative 
influences so that two species solidly maintain themselves, differs because radiolata 
is more susceptible to decimating conditions and is consequently compelled to 
multiply at a slightly higher rate. Which of the foregoing suggestions is correct is 
difficult to determine without making a similar study of fertility at some other 
time of the year. 
TABLE 43.—Partula gibba, Guam. Apra, West Central, and Southwest Regions. 

Lip. Tooth. 
Series. No. 
Yellow; 
White. yellow- None. | Trace. 
brown. 
Orote: mii tel lan oy, xaos Bhs 19 3 16 19 
Agat: sitellacy Fass Oe 6 ces 6 6 Bed 
Salifan: bicolor sacs s.< 5% i tc 1 1 aie 
mitetla. 2625: br 85 85 76 9 
Umatac Salonga: bicolor......... 7 ae Ps 7 
Umatac Pajom: bicolor......... 9 2 7 9 
Umatac Madog: bicolor......... < f 7 7 
Merizo: bicdlor.. < 5:>.« ae 1 1 1 

Inquiring more closely into the existence of a defined breeding season, we may 
deal with the problem as in the case of radiolata._ Using the data given in table 45, 
correlations may be worked out between the passage of the weeks and the number 
and nature of the embryonic contents, employing the 24 collections which are 
sufficiently numerous for the purpose. The result is zero for the correlation 
between the lapse of time and the percentage of gravid individuals which means 
that certain snails conclude their reproductive phase as rapidly as others enter it. 
When the lapse of time and the total contents are correlated, the figure is —0.0230 
+0.1376; and when the passage of the weeks is correlated with the number of 
eggs, the result is —0.0600+0.1372. Both coefficients are less than their probable 
errors and therefore are not intrinsically significant. 
TABLE 44.—Partula gibba, Guam. Statistics of tooth development in the several color-classes. 



Number. Per cent. 
Color-class. Total 
No. 
None. Trace. | Small. 
URICONGT SoBe et nee nas okt har 150 110 34 6 1368 228i. 4.0 
Bicalor ss oct. oo0 bc. dene a ois 164 162 1 1 98.8 0.6 0.6 
mitella 220.5,0 85 sper tse een 1,309 LT, T5t 139 19 87.9 10.6 1.4 
mitellacrabraic, «foes «Ga clon «eras 264 208 34 22 78.8 12.9 8.3 
fo RSS NE RIA Ae ee SWS it ee 245 178 56 11 72.6 22.8 4.5 
castaneto3a) Coc nee i: dae esa des 46 45 1 iwol 97.8 22 Mee 
WESDEIR Ron So nla CER 220 168 43 9 76.4 19.5 4.1 
marginatascd ss: ceeds dads oe eyes e 6 4 2 hes 66.6 Sone oar 
SumMmMary fecis pone hee Tee 2,404 2,026 310 68 84.2 12:9 2.8 
However, such results would follow if a definite reproductive period did occur in 
gibba and began earlier than in radiolata. Then the total contents would naturally 
