VARIATION AND ZOOGEOGRAPHY OF LIGUUS IN FLORIDA. 435 



one may show a strong tendency to restrict a given pattern, another to spread 

 it (cf. PL XXXIX, figs. 20g-20j, Miami, and figs. 22o-22d, Cutler). 



Each character acts as a unit in variations of potenc . Thus in the k 

 north of Largo, pink color is unusually strong in several colonies, but yellow is 

 very strong in some, almost absent in others. The green lines also vary ind< 

 pendent ly of yellow or pink. Moreover, the loss of markings of one color does 

 not affect the rest of the pattern. 



Hybrid colonies of Liguus are found only in places where the topographic 

 conditions have permitted two or more pure races to mingle in the same wood- 

 lands. That the various forms of these colonies actually hybridist is shown 

 by the fact that the eggs laid by one parent produce several color-forms. Of 

 four sets of eggs on the point of hatching examined from the Miami colony, thr 

 contained embryos of three patterns, the other of two patterns. One set of 

 nine eggs deposited in the earth under a bell-jar by a snail rought by me from 

 Miami, similar to PI. XXXIX, fig. 20k (roseatus pattern), male parent unknown, 

 hatched into 4 crenatus pattern, 2 roseatus pattern, 2 casta neommatu & patn-rn, 



1 egg not broken. 6 



In all hybrid colonies there is almost complete segregation of the component 



patterns. The only evidence of blended inheritam een is in certain 





< 



very rare individuals, apparently crenatus X roseatus, from outli of Miami 

 which the apex is white but the parietal wall shows faint pink coloring. Possibh 

 those castaneozonatus in which the chestnut color is reduced to a few pale sjwts 

 (as in PL XXXIX, fig. 20;) might be considered as blends, but they admit of 



another explanation. 



The form castaneozonatus is nowhere known as a pure race It is likely that 

 roseatus arose from (gradual?) loss of the ability to deposit brown color in some 

 one colony, which later attained a wide distribution; subsequent restoration of 

 communication with the parent stock of castaneozonatus resulted in a wide- 

 spread hybrid race. 



Incipient races formed by the appearance of new characters, probably by 

 mutation, are seen in L. solidus pictus, L. fasciatus marmoraius and L. /. V tu- 

 dineus. The dark pattern in these is not a modification of the ordinary pattern, 

 due to diminishing or increasing potency in its ontogeny (as are the various 

 local forms of L. /. castaneozonatus, etc.). It is a new pattern. In fig. 2, it will 

 be seen that in the normal patterns such as castaneozonatus (fig. 2, A) and hg- 

 numvitce (fig. 2, B) there is a dark border (s) immediately below the suture, 

 followed by a white band ; at the periphery there is a dark line (p) bisecting a white 

 band. In testudineus (fig. 2, C) the sutural and peripheral lines («', pT I are white 

 and the positions of the white bands of the other forms are occupied by inter- 

 rupted dark bands connected with the longitudinal flames. They are negatives 

 of the normal patterns. The embryonic and early neanic whorls never have the 



• 



ems are readily distinguishable in the embryo snails by the presence of clear roje 

 rose color with darker stripes in castaneozonatus. I do not know that crenatus couia 



be distinguished from testudineus at this stage, both being 



