EVOLUTION OF FOSSIL CEPHALOPODA. 247 
ants, and by a comparative study of larval stages with 
adult forms the naturalist finds the key to relationships, 
and is enabled to arrange genera in genetic series. They 
were all marine, never parasitic, always free, and so with 
them there is no obscuring of the record; also in the 
Mollusca generic and specific characters show in the 
shell better than in the soft parts; so the classification 
of fossil ammonites is just as good as that of living 
shellfish. 
The importance of this will be appreciated if we con- 
sider other groups. The crustacean goes through suc- 
cessive stages and moults its shell after each change, 
thus making it difficult, especially in fossils, to find and 
study these various stages. The ammonite, after it has 
grown out of each stage of its shell, carries it around 
enveloped in the later chambers. So in a single speci- 
men the record is complete, and may be read by skilful 
handling. 
Although genera appeared in the order of corre- 
sponding larval stages, they did not disappear in the 
same order; and so their survival under favourable con- 
ditions is liable to make confusion in the record, if one 
depends wholly on the study of series of adults. Such 
forms, for instance, as Styrites, Tropiceltites, Miltites, and 
others, that are now known only in the Karnic zone of 
the Upper Trias, are undoubtedly such survivals, for 
they still have simple goniatitic sutures, very little orna- 
mentation, and in general are more like Lower Triassic 
ammonites than members of the TZvofites subbullatus 
fauna. The stray Ztrolites foliaceus, which appears in 
the Alps and in California in this same fauna, is another 
survival of a Lower Triassic type, but fortunately we 
do know TZ?rol/ites in the horizon where it belongs. If 
this were not the case the naturalist would be very much 
puzzled at finding Zvrachyceras of the Karnic horizon 
