189
Vol. 36(2), 2001
(Linne) (Taylor & Miller, 1989:230, figs 4-6,22b) and other cerithids (Risbec, 1943). The long, monopectinate osphradia of Bittium, Finella and Alaba were here included in the same group as the Cerithium osphradia; the unipectinate condition is regarded as a result of miniaturization, similar to what has occurred in Columbellidae among Muricoidea (person, obs.) and Pisanianurinae among Ranellidae (Waren & Bouchet, 1990). The second type of pectinate osphradium is that of Campanile. This is similar to those of higher caenogastropods, and has caused some confusion in the systematic allocation of this species. In fact, the Campanile osphradium superficially resembles those of neogastropods in having an elliptical outline, but careful analysis shows that it differs greatly. Contrary to those of higher caenogastropods, the osphradium of Campanile has (fig. 422): 1) the leaflets inserted in the mantle in the entire superior region (while in the higher caenogastropods the leaflets are free, mainly inserted near osphradial ganglion and adjacent area); 2) a proportionally greater number of leaflets, being apparently thinner and morę densely packed; and 3) the insertion of the right leaflets in ctenidial vein which, observed intemally, apparently presents eflfective action in blood osphradium drainage. The former character is shared with the pectinate osphradium of the Architaenioglossa (person, obs. aaPomacea spp.). In the present analysis, both the pectinate cerithioidean osphradium groups, with elear differences, are allocated in different points in the tree, suggesting independent origin. An almost certain independent origin of the pectinate condition of the osphradium is also found in higher caenogastropods and ampullariids. Bipectinate osphradium is also known in other cerithioidean taxon, Plesiotrochus (Houbrick, 1990b). However, the affinity of this taxon with those above mentioned is matter still unclear.
29. Antcrior extremity: 0= posterior to gili; 1= ante-
rior to gili (Supplanaxis nucleus, Alaba incer-
ta, Serpulorbis decussatus) (CI= 33, RI= 0).
30. Length: 0= about 2/3 of gili; 1= less than 1/2
gili (Aylacostoma ci, Melanoides
tuberculatus, pleurocerids, Cerithidea
costata, Campanile symbolicum)-, 2=
equivalent to gili (Aylacostoma ezoplicata, Supplanazis nucleus, Turritella hookeri, M. modulus, cerithids, Finella dubia) (CI= 22, RI= 46, not additive).
The length of the osphradium relative to that of gili, appears to be an important character at the specific level, but poor for analysis at higher ranks, because it may vary within a single genus (e.g., Aylacostoma) and appears in several points in the tree (see Iow indices). Another character of poor value for higher systematics, for the same reasons, is the position of the anterior osphradium end in relation to that of the gili (if anterior or posterior to it).
31. Anterior extremity shape: 0= simple; 1 = strong
zigzag (A/, modulus)-, 2~ delicate zigzag (Doryssa atra, D. macapa) (CI= 100, RI= 100, not additive).
The osphradium with a zigzag in its anterior region appears to be another adaptation for inereasing the surface of this sensory organ as an altemative to the pectinate condition. This is morę developed in Modulus.
32. Distance from the left margin of the pallial
cavity: 0= close; 1= far (Cerithium atra tum, Alaba incerta, Serpulorbis decussatus) (CI= 33, RI= 0).
33. Satellite fold: 0= absent; 1—at left (Aylacostoma
exoplicata)\ 2= both sides (Bittium varium, Finella dubia, Alaba incerta) (Cl= 100, RI= 100, not additive).
Two types of satellite osphradial folds, apparently of independent origin, were found, a type in Aylacostoma exoplicata, restrict to the left side of osphradium, located between it and left margin of the pallial cavity. Other type occurs in miniaturized forms (Bittium, Finella and Alaba) surrounding almost entire the osphradium length (except for the posterior region) and appeared as a shared synapomorphy. A similar fold was also described for Faunus ater (Linne) (cf. Houbrick, 199 lb) but sited at right of osphradium, between the later and the ctenidial vein (Houbrick, 1991 b, fig. 23B: glandular ridge). The satellite fold resembles a little the endostyle present in filter feeding gastropods such as Crepidulidae and Struthiolariidae (person, obs.)