46 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
it continues through adolescence to maturity; therefore it constitutes a real color- 
character, worthy of full recognition. Another matter of interest is the wide dis- 
persal of such shells in places as far separated as Cabras, Umatac, and the East 
Central Region; while this point is reserved for future discussion, it is sufficiently 
striking to be emphasized here. 
The sixth and last color-class is distinguished as bicestata (figs. 49 to 56, plate 11). 
The most typical shells are as light as flavea or even pallida in ground-color; revolving 
purplish or purplish-brown bands lie along the sutures of the upper whorls, and 
extend“over the body-whorl with greater or lesser distinctness nearly to the lip. 
In the Salifan locality the deepest colored shells are found (figs. 49 to 51, plate 11), 
while in the Apra and the northern regions the colors are lighter, the revolving bands 
are browner, and the bands more frequently fade out as they extend over the last 
whorl (figs. 52 to 56, plate 11). ‘There is no possibility of confusing this type with 
strigata-helix, even though the apical coloration is similar. Here too it must be 
noted that adolescent shells are found which display the typical markings (figs. 
55,56, plate11), and even many of the embryonic shells also exhibit the apical helix 
and the revolving bands; hence there is no question of changing color during 
growth. Light examples of this class are undoubtedly what Pilsbry had in hand 
in describing ‘“‘a form collected by Quadras’’; in fact, the author has seen the par- 
ticular shell figured by Pilsbry (Manual, vol. 20, fig. 2, plate 41), and it is exactly 
like specimens secured personally in the Agafia area, in which the bands of the 
last whorl are evanescent, although the markings of the spire and the general 
ground-color stamp the shell as a strigata-helix. 
Such, then, are the distinguishable color-classes of radiolata. If we take them 
on the basis of their manifest qualities, their inter-relationships are capable of 
representation as in text-figure 3 already mentioned, which is self-explanatory. 
It is quite another question as to what the actual genetic relations may be, and 
there is nothing in the material which would justify a final statement in factorial 
terms. Nevertheless, it is permissible at least to suggest the factor basis. If the 
typical yellow factor of flavea were diminished in strength, or were entirely absent, 
pallida would result; if it were supplemented by the factor for reddish, fulva would 
arise. [he factor for transverse streaks is universally lacking in the foregoing 
classes, but when it is present it is accountable for the type of coloration called 
strigata; when this is accompanied by a determiner for revolving bands, the result 
is the strigata-helix pattern. In the absence of the factor for strigation, and with 
the extension of the effect of the last-named factor for revolving bands, the bicestata 
type would result. The actual relations, to repeat, may be different, but for pur- 
poses of summary description the terms of factorial elements have a certain utility. 
Reverting now to the phenomena of distribution, it is entirely clear from the 
census table (table 16) just how the colonial color-composition varies from place to 
place and from region to region. The lighter color-classes of flavea and pallida are 
almost ubiquitous, while the others occur only sporadically. The dominance of the 
strigated shells in the Apra territory is a striking feature. The scattering occurrence 
of fulva is another notable point. Still more interesting is the existence of bicestata 
