OE SOUTHERN INDIA. 245 



The hinge varies greatly ; in typical forms there are two cardinal teeth fitting 

 alternately one behiucl the other, and one lateral tooth on each side ; the ligament 

 is external or sub-internal. In other closely allied forms either the laterals or the 

 cardinals become partially or wholly obsolete, and again in others, all teeth are 

 absent dm-iug all stages of age. 



It appears to me that two groups or sub-families could conveniently be 

 separated, though their definition is somewhat arbitrary. The first, corbin/e, has 

 as type the genus Cordis; the shells are generally elongately ovate, or ovately 

 rounded, solid, with muscular impressions sub-equal, broadly ovate, the ligament 

 always external. The second, lucininm, of which the type is Lucina, have a 

 more or less orbicular shell, the anterior muscular impression is narrower and 

 much longer than the posterior, the ligament is lodged in a deep groove or is 

 sometimes nearly internal. As these two divisions appear also convenient for the 

 grouping of the numerous fossil forms I shall here retain them. The following 

 genera have to be classed in this family ; (a), UiucanUwm, Conchocele, Fimhriella, 

 Gonodon, Sphcera, Ilutiella, Spluvriola, Corbls, Corbicella, Sportella, Sphce reiki ; 

 fbj, PhlUs, Cri/ptodoii, Lucina, (including the sub-genera Lorlpes, Sere, Codakia, 

 Mlltha, Ilyrtea, Cyclas). 



Eor the palaeontologist the study of this family is of great importance, (there 

 are several forms which may prove to be characteristic types), and to the concho- 

 logist its development is very interesting. The Lucisid^ make their first appear- 

 ance already in siluiian rocks, though the species agree with the type only in 

 external shape. I am not aware that the determination of any of the palaeozoic 

 forms has as yet been authenticated by an examination of the hinge, but in 

 most instances the species seem to possess great claim to accuracy of determina- 

 tion. In the Trias we have a few species of the combine, some forms interme- 

 diate between both sub-families, like Sph(era and Uiiioardinm, prevailing ; the 

 LUciNiNM are also represented, but not yet well discriminated. In the Jurassic 

 period the number of species of both sub-families gradually increases from the 

 Lias upwards. The two groups which I have pointed out as sub-families ap- 

 pear to become clearly indicated. The corbin^ are decidedly more numerous 

 and varied in form, the genera Zfnlcardlum and Splioiriola and probably Sphcerella 

 attain their highest development ; typical Corbls, and most likely also Flmbrlella 

 and 3IiUiella, are also found ; Corbicella is a new form, which again disappears 

 with the beginning of the cretaceous period in which SphcBra is added to the 

 number. As to variety of generic types, and perhaps even as to species, the two 

 periods, Jurassic and cretaceous, are nearly equal, but almost all these forms seem to 

 disappear in the tertiary period, where we find Conchocele and Sportella, particularly 

 the latter, very numerous ; it is the representative of Corbicella. Corbls alone con- 

 tinues up to the present time, but the number of species decidedly decreases. 



In the LuciNiNJE the species gradually increase in number from the Lias, 

 through the Dogger, Malm, and the whole of the cretaceous period. Some of the 

 elongated sub -genera, like Ilyrtea, which in form approach the combine, are here 



