26 INTRODUCTION. 



perhaps a little of the lime. Second, cases in which there has 

 been a more or less complete replacement of the original com- 

 position by some foreign material — as iron pyrite or silica, for 

 example — ^through a process of infiltration. lu these two in- 

 stances the original structure of the tissues remains intact to 

 a greater or less extent. Third, those in which the hard parts 

 have been entirely removed, and the cavities thus left in the 

 rock are completely tilled with silica or some other substance — 

 the external characters being commonly as well preserved as 

 when infiltration takes place, but no microscopic structure is 

 ever apparent. Fourth, a set of conditions much like the last, 

 but the cavities not again filled by foreign minerals. Moulds 

 of the outer surface frequently show the external characters 

 and ornamentation perfectly; and good reproductions of the 

 original forms may be easily made by taking wax, or gutta percha, 

 or even plaster casts. Fifth, in which the remains are known 

 only from internal casts, as among many gasteropods and 

 lamellibranchs. The fine mud in which the organism was 

 finally buried worked its way into the interior of the shell, 

 eventually forming a compact cast of the inside. After the 

 deposit had hardened into rock, the shell itself was dissolved 

 away, leaving the inside impression intact. In many cases both 

 the internal cast and the exterior mould are found together ; 

 but oftener when the cast occurs the outer impression is not 

 at all perfect, and when the mould is good the internal cast is 

 often composed merely of loose sand or clayey material, which 

 falls into the bottom of the cavity when the hard parts pass 

 away in solution. Sixth, when the evidences of life are in mere 

 traces or indifferent indications. While these have no special 

 value, either morphologically or stratigraphically, they serve to 

 prove the existence of particular groups at certain horizons ; 

 and often they lead to the discovery of more important re- 

 mains. 



It is manifest that only the first three categories mentioned 

 are of special use in detailed faunal considerations. The first 

 and second kinds are the most satisfactory of all; but the third 

 and fourth furnish many suggestive hints, particularly when 



