PARTULA HYALINA. 53 
While the matter of length is unimportant, the diversity in width, and therefore in 
shell proportion, is decidedly indicative of real island differentia. As regards the 
aperture, no profound differences are found, except in the case of its length as 
compared with that of the whole shell. In all seven characters the Moki shells are 
less variable, and significantly so in four cases. 
TaBLe 16.—Partula hyalina. Cook Islands. Statistical description. 
Mean VALUE. 
Shell. Aperture. Wenethtaperture 
Locality. No. + length shell, 
Length. Width. Proportions. Length. Width. Proportions. ID RODONO EL 
mm. mm. p. ct. mm. mm. p. ct. p. ct. 
Mangaia, dead. .| 46 | 15.9174+.0814) 8.9565 .0295| 56.1087 +.1826| 9.3043 + .0304) 6.6667 = .0933| 71.5667 +.2221) 58.4783+.1766 
Moki, dead..... 38 | 14.9553+.0817| 8.3789+.0390) 55.8947 =.1870) 8.5621+.0440) 6.2842 + .0331) 73.2632 .2405| 57.1053 + .1637 
Moki, living....| 15 | 15.4700+.0970| 8.4067 +.0400) 54.3667 +.2192| 8.8667+.0550) 6.4533 .0371| 72.8333 .2962| 57.2323+.2155 
Moki, all....... 53 | 15.1001+.0701| 8.3868+.0289| 55.4623+.1608| 8.6340+.0377| 6.3321+.0269} 73.1415+.1853) 57.1415 + .1323 
STANDARD DEVIATION. 
Mangaia, dead..| 46 | 0.8185+.0814) 0.2968+.0209| 1.8353.1290} 0.3056+.0215)} 0.9261 .0658} 2.2060+.1568] 1.7753 =.1248 
Moki, dead..... 38 -7476+.0579| .3562+.0276| 1.7097=.1323) .4023=.0311| .3030=.0235| 2.1983+.1701|) 1.4963+=.1158 
Moki, living....} 15 -5564+.0686| .2294+.0283) 1.2578=.1551} .3155+.0389) .2126+=.0262 1.6996 +. 2096 1.2364 .1524 
Moki, all....... 53 .7575+.0496| .3120+.0204| 1.7370+.1138| .4069+.0266) .2906+.0190 2.0008 1311 1.4287 + .0936 
TaBLe 17.—Partula hyalina. ‘“‘Living” shells of Moki as compared with those of Tahiti. 
Character. Mean value. Standard deviation. 
Shell Mien oth mimes sclerosis tie isi ee screen ent [+0.1602= .1000]|[—0.1238=+ .0707] 
wid tbh wim miss cisleys Suen eine eislesd acu Acre oes — .4348+ .0420 | — .1862+ .0297 
DY KOV LOANS, fe GAME ocogosc0c00n00b000000 —2.6560= .2286 | — .7109+.1615 
Aperture sieng th sin meee anEnnernrrrnericnnicnticcc [— .0357+.0570]| — .1627+ .0402 
Wid CHW A SER cree heey ee Oe [— .0495+.0390]| — .1726+ .0278 
DIRODLOFANLOVER FI? GAM 5000000000000 00c06 [— .1181= .3048]|[— .6038+ .2155] 
Length aperture + length shell, proportions, per cent.) —0.8055= .2155 |[— .4129+ .1566] 
Of the 15 adult hyalina taken on Moki, 14 proved to be gravid (93.3 per cent). 
It was still more surprising to find that these parents yielded 31 eggs and 72 young, 
giving a total of 103. Thus the average number of embryonic individuals is 7.35 
for gravid adults and 6.86 for the whole series. These figures indicate a much 
higher degree of fecundity than in the case of Tahitian snails and, if we should treat 
them as we did the figures in the case of the latter, it would seem that elimination 
is far more searching in its effects on colonies living on relatively low islands com- 
posed of coral limestone. Only 1 in 7 new-born young is able to reach reproductive 
maturity. However, we can not place the same reliance upon these figures as upon 
the fuller data relating to the snails of Tahiti. 
