94 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
The statistics of heredity (table 39) are entirely trustworthy in the case of the 
coil; each form breeds true, as far as the meager figures go. In the matter of 
color, however, the observations are less significant, owing to the weak develop- 
ment of ground-color and color-pattern until the young snail is nearly ready to 
leave its parent. In the immature condition it is impossible always to discriminate 
between the two classes of unbanded, or between the two groups of banded shells. 
Nevertheless, there are enough positive data to prove that plain parents produce 
banded young, and vice versa; hence the several classes of this variety are not so 
greatly differentiated as to be physiologically discontinuous. 
Partula nodosa composita var. nov.—Punaruu Valley. 
The topographical features of this valley—the original locality designated by 
Garrett—have been described and illustrated on earlier pages. Collections were 
made in 1906, 1908, and 1909, from substantially the same segment of the valley. 
The snails agree closely with those of Taapuna, to which, indeed, they stand in an 
ancestral relation; hence by prior right they bear the same varietal name, composita. 
The first specimens of nodosa were taken at about 13 miles from the coast, at 
a height of only 125 feet above the sea-level. From this point inward they increased 
in numbers, occurring on both sides of the stream, and not from the south side solely, 
where Garrett obtained them. At a little over 3 miles inland, nodosa disappeared 
and only P. otaheitana was taken; the barometric level at the limit of nodosa’s 
range was 350 feet. Since Garrett’s time, therefore, this species has come to occupy 
both sides of the valley floor, for a radial distance in excess of the 0.75 mile specified 
by Garrett. 
Tasie 40.—Partula nodosa composita. Punaruu Valley. 
= Class I. Class II. Class III. Class IV. 
No. of 
No. of 
Year. ASHI, adoles- 7 } 
cents. | 4 quits. Adoles-|Per cent menial Adoles-|Per cent) anita Adoles-|Per cent DGHIRS. Adoles-|Per cent 
| cents. | adults. cents. | adults. cents. | adults. cents. | adults. 
IMOOGs oao0 73 48 17 Uf 23.3 50 37 G35 |} . § 3 6.8 1 i 1.3 
190 See 35 6 6 1 7/61 28 5 80.0; 1 (0) 2.8 0 0 0 
MNO). soo 78 50 42 24 53.8 36 25 46.1} 0 1 0 0 0 0 
IAN oreo: ete 186 104 65 32 34.9 114 67 61.3 6 4 357 1 1 0.5 
| 
Although nodosa inhabits a restricted and uninterrupted region of the valley, 
the three annual series from this area are by no means the same as regards propor- 
tionate numbers in the four color-classes. ‘The specific facts are given in table 4o. 
As compared with the colony of Taapuna, the Punaruu group as a whole includes 
more individuals of Class I, while Classes III and IV are very poorly represented. It 
is certainly noteworthy that only one example of the last group should be found when 
more than one-third of the 1909 shells of Taapuna belong to this class. The shells 
of Class I are more strongly strigated than in Taapuna (figs. 1 to 4, plate 23). The 
white sutural area is more prominently displayed in the banded shells of Class I 
(figs. 5 to 11, plate 23) as well as in those of Classes III and IV (figs. 12 and 13, plate 
23), where it is more sharply marked by contrast with the general dark ground-color. 
