122 THE NAUTILUS. 



shoulder of the whorls and in the accentuation of the three 

 principal carinas. 



Philippi's figures of tricarinatus copied on pi. VIII, figs. 2-3, 

 are not quite so much elevated and look not unlike V. javanicus 

 luzonicus as figured by Kobelt (pi. 46, fig. 9) and herein. 

 Philippi notes several minor differences between his specimens 

 and Anton's description, but "has no doubt" but that his 

 shells are correctly identified. 



In the absence of specimens with their opercula, the standing 

 of Anton's species and its relations to both costatus and javanicus 

 must remain uncertain. 



XII. 

 VIVIPARUS JAVANICUS LUZONICUS Kobelt. PI. IX, fig. 8. 



Vivipara javanica luzonica Kobelt. Con. Cab., Viviparida;, 

 1909, p. 378, pi. 46, figs. 5, 9 and 10. 



The Sarasins (Suessw. Moll. Celebes, 1898, p. 59) were the 

 first to call attention to the radical difference in the opercula of 

 the two species, V. costatus Q. and G. and V. javanicus v. d. 

 Busch, which in their shell characters are often quite indis- 

 tinguishable. 



In javanicus and its allies the central part of the inner side of 

 the operculum is occupied by a granulated area, which is sur- 

 rounded by a smooth, polished border. 



In costatus, on the other hand, the central portion is smooth 

 and polished, but is surrounded by a narrow, distinctly granu- 

 lated area and outside of this the remainder of the surface is 

 smooth and polished like the centre. 



Kobelt (1. c. ) has described a race from Daraga, Luzon, 

 which has the typical javanicus operculum, but in other re- 

 spects closely resembles costatus. I have similar specimens with 

 their opercula, figured above, which were collected in the Phil- 

 ippines by Steere, but no exact locality is given. 



Among the shells received from Mr. Webb was a single speci- 

 men from Panique, Tarlac Prov., Luzon, which agrees in its 

 shell characters with the Steere specimens, but unfortunately 

 has no operculum. 



If Anton's tricarinata should prove to be identical with this 

 form, his name would have priority. 



