408 NEW GENUS AND SOME NEW SPECIES 



strongly characteristic, but we cannot place much reliance on 

 this character, unless accompanied by and dependent on others. 

 Thus, the angle of the cardinal tooth depends much on the 

 location of the beaks, and we know that in the same species 

 the location is quite different, and yet this difference is not 

 worthy of creating even a variety. If the beaks be placed 

 immediately over the anterior margin*, as in the ellipsis they 

 generally are, then the cardinal tooth will be nearly or quite 

 parallel to the lamellar one ; but if the beaks be more posteriorly 

 placed, then the cardinal tooth becomes more oblique. We 

 must, therefore, when characters are so difficult, look at them 

 in combination, and adopt them with due consideration. 



In the same species the mass or substance of the valves 

 varies much according to localities. Thus we find the com- 

 planatus\ in some of our Atlantic rivers full grown, when 

 only an inch broad, while in other of our Atlantic rivers 

 we have them four inches broad. In some localities we 

 have them possessing but little calcareous matter, while in 

 others they are almost massive. This also occurs in per- 

 haps a greater degree with some of the western shells. And 

 if we examine a massive specimen, we are almost sure to find 

 the cardinal teeth more or less thick, whilst those of the same 

 species which are thin, and they frequently differ very 

 much in this respect, will be found to possess cardinal teeth 

 of quite a crested structure. The cardinal tooth, being single 

 in one and double in the other valve, or double in both valves, 

 cannot be depended on as an unfailing character. The same 

 species will often present double teeth in both valves, although 

 it may be usual to possess them in the right valve only. The 

 lamellar tooth depends much on the substance of the speci- 

 men. If it be massive the teeth will be thick, if thin more 

 bladed ; the teeth, therefore, differ almost as much in varieties 

 as in species. We must, consequently, while examining a 

 specimen to determine its species, give due attention to these 

 counteracting characters. 



* I reverse Lamarck's anterior and adopt Cuvier's as herelofore. 

 t Purpurcus of Say. 



