CHAP. V. ROSTELLARIA. APORRHAIS. 141 



ascends from the aperture nearly to the top of the spire : 

 the shell has much of the shape of a typical Pleurotoma ; 

 the basal channel being more or less lengthened. To this 

 genus we are led by the Stromhidea tridentata, — the only 

 Stromhus Avhich gives us at once the outer lip of Ros- 

 tellaria cwvirostris, with all the other characters of our 

 genus Stromhidea, The first type of Rostellaria is seen 

 in those elegant shells, R. curvirostris and rectirostris, 

 where the siphon is moderate : the outer lip is hardly 

 dilated, since its margin is merely divided into short 

 processes. The rectirostris is of great rarity ; but of 

 still greater value is that which forms the second type ; 

 indeed, it seems to be known only by the figure of Chem- 

 nitz, and has been altogether omitted by Lamarck : it 

 is channeled to the tip of the spire ; while the outer lip, 

 without being much dilated, is divided into a number 

 of short spine-like processes. In the third type, Ros- 

 tellaria macroptera, the upper siphon is equally long ; 

 but the lip, instead of being cleft, is entire, semicircular, 

 and dilated to an enormous size. In these two last tvpes, 

 in fact, we have the greatest developement of the lip; and 

 hence we may expect a declension in the succeeding forms. 

 Such is accordingly seen in Rostellaria columbata, where 

 the lip, although dilated and very broad, does not extend 

 more than half way up the spire, although it is rather 

 lengthened into the same sort of lobe as that in Strombus 

 pacificus. There is still a great difference between R. 

 columbata and the first type we noticed ; but our fifth 

 brings them together : this is formed of the small species, 

 four or five in number, among which the elegant little 

 Grignon fossil, Rostellaria fissurella Lam., is the best 

 known. In these shells, the lip is no longer dilated, nor 

 is it extended up the spire ; they lose, in short, one half 

 of the characters of their own genus, and in the same pro- 

 portion assume those of Strombus, — thus becoming the 

 most aberrant type of the sub-family; while their station 

 between R. curvirostris and columbata is apparent to the 

 most unscientific eye. 



(131.) The last genus is that of I)a Costa's Apor- 



