CONCHIFERA. 405 
~ 
the shells could not be determined from such specimens. Our 
knowledge of the extinct Radiolite is derived from natural 
moulds of the interior, formed before the dissolution of the 
inner layer of shell, or from specimens in which this layer is 
replaced by spar. 
The necessities of geologists have compelled them to pay very 
minute attention to the markings in the interior of shells, to 
their microscopic texture, and every other available source of 
comparison and distinction. It must not, however, be expected 
that the entire structure and affinities of molluscous animals 
can be predicated from the examination of an internal mould or 
a morsel of shell, any more than that the form and habits of an 
extinct quadruped can be inferred from a solitary tooth or the 
fragment of a bone.* 
The systematic arrangement of the bivalves now employed is 
essentially that of Lamarck, modified, however, by many recent 
observations. The families follow each other according to rela- 
tionship, and not according to absolute rank; the Veneridce are 
the highest organised, and from this culminating point the 
stream of affinities takes two courses, one towards the Myas, 
the other in the direction of the oysters; groups analogically 
reiated to the Tunicaries and Brachiopoda. 
SECTION A. ASIPHONIDA. 
a. Pallial line simple: Integro-pallialia. 
Tam. 1. Ostreidee. 4, Arcadee. 
2. Aviculidee. 5. Trigoniadee. 
3. Mytilidee. 6. Unionidee. 
SECTION B. SIPHONIDA. 
7. Chamidee. 11. Lucinidee. 
8. Hippuritidee. 12. Cycladidee. 
9. Tridacnidee. 13. Cyprinide. 
10. Cardiadee. 
b. Pallial line sinwated: Sinu-pallalia. 
14. Veneride. 18. Myacide. 
15. Mactride. 19. Anatinidee. 
16. Tellinidee. 20. Gastrocheenide. 
17. Solenidee. 21. Pholadidee. 
The characters which have been most relied on for distin- 
* Etudes Critiques sur les Mollusques Fossiles, par L, Agassiz, Neuchatel, 1840, 
