PARTULA PRODUCTA. 303 
The even dark-brown color of some specimens (plate 34, fig. 63) is the product 
of an ultimate suffusion of bands that are elsewhere more or less vague but recog- 
nizable (plate 34, figs. 61 and 62). Finally, very light specimens occur (plate 34, 
fig. 64.) which seem to be derivates of faintly banded types like that of figure 60, 
plate 34, in which the stripes have become evanescent; in practically every case, 
in my own and in other collections, some trace of the bands is discoverable, even 
though it may be very faint. The point at issue is that the three color-types of 
sinistrorsa, namely apex, cestata, and phea solida are not repeated exactly in this 
species; the three forms actually found are reducible to one, which does not agree 
with cestata in the details of the revolving bands, however much it may resemble 
the latter on casual inspection and acquaintance. 
Defined statistically (table 252) the species proves to be somewhat similar to 
its sinistrorsa associates, but it is clearly distinct, especially in the characters of 
the aperture. Only one shell exhibited a weakly developed trace of a pillar tooth. 
TaBLE 252.—Partula producta, Faarahi Valley. 
Statistical description. 
MEAN VALUE. 
Shell. Aperture. Length aper- 
5 ture + length 
Series. | No. 
shell, propor- 
Length. | Width. Proportions. Length. Width. Proportions. tions. 
— 
| mm. mm. p. ct. mm. mm. p. ct. p. ct. 
Live...| 29 | 18.9224 .1180) 10.2172+.0403) 53.9482+ .2474| 9.5758 .0571) 7.0449 .0308] 73.3965 .2948} 50.4311+.1971 
Dead..| 14 | 19.2142 .1503) 10.1427=+.0736) 52.9286+ .4899) 9.6143=+.0814| 6.9000+.0721) 71.9286+ .6045| 49.9286 .2785 
2 = 
All. .| 43 | 19.0174 .0943) 10.1930+ .0364| 53.6163+.2360) 9.5883 .0468) 6.9976+ .0321| 72.9186 .2886| 50.2675+.1628 
STANDARD DevIaTION. 
Live. 29 | 0.9428+.0834) 0.3217+.0285| 1.9753+.1749| 0.4561+.0404) 0.2457 .0218) 2.3539 .2084| 1.5741+.1393 
Dead..| 14 -8338+.1063) .4083+.0520! 2.7180+.3464| .4517+.0575| .4000+.0510) 3.3533+.4274| 1.5452+.1969 
All. .| 43 9172 .0667| .3539=£.0257| 2.2947=.1669|~ .4551=£.0331) .3122+.0227 28056-2044 1.5824 .1151 
The data on fecundity are: records, 29; gravid, 25; per cent gravid, 86.2; 
eggs, 26; young, 17; total contents, 43; average for gravid, 1.72; average for 
all, 1.48. 
Of the young, 16 were dextral and banded; only 1 was devoid of recognizable 
bands, but its whorls were lighter just below the suture, so that its condition sup- 
ports the view expressed above as to the relationship between the light plain shells, 
and the banded individuals; this embryo, like the rest, was dextral. 
As Partula producta was found in one situation only, it does not afford an 
opportunity for the study of distribution. Its intrinsic characters and internal 
variation are of some interest, especially in connection with a wider analysis of the 
whole genus, when terrestrial species of other islands, notably Raiatea, are investi- 
gated. For the present, it stands alone on account of its many points of contrast 
with the other forms of Tahiti. 
