76 GENESIS OF THE ARIETIDA. 
in any other species of this genus. Thus a high degree of specialization in the 
development of keel, channels, and pils is correlative with decidedly retrogressive 
changes. 
In Coroniceras, the Bucklandi series exhibits very decided changes in both 
the individuals and the species. The tubercles were first lost during the old age 
of the individual, the sides became more convergent even in Bucklandi itself, the 
abdomen narrower, the pile reduced to folds and bent like those of the adult of 
Ast. oblusum, the channels shallow and finally almost obsolete, and the keel, even 
though becoming apparently more prominent on account of the convergence of 
the sides and obsolescence of the channels, was really not so sharp or well 
defined. 
Cor. orbiculatum exaggerates all these old age changes, becoming narrower on 
the abdomen, with more convergent sides, and this convergency began even 
in the ephebolic period in some examples. Similar changes occurred very late 
in the life of individuals in the next subseries. Thus even the convergency of 
the sides was not found in the adults of Cor. rotiforme in many specimens, and is 
but slightly developed even in the extreme old age of some of this species, and in 
its predecessor, Cor. kridion. 'This characteristic is, however, observable habitually 
in the adults of Cor. ra. These lead into Cor. Gmuendense of the same series, 
which had very convergent sides in the adult, and was often also destitute of 
tubercles. The last were confined to the earlier stages of this species, and in old 
age the changes were very marked and rapid. The extreme variety of Cor. 
trigonatum inherited convergent sides, smooth and half obsolete pile, narrow 
abdomen, shallow channels, and elevated keel, so early that we may say with 
confidence they all appeared in the ephebolic period. 
The old whorls of Cor. G'muendense and Cor. trigonatum™ have the sides of the 
whorls convergent and a decidedly trigonal form. This form is correlated with 
obsolescing pile and a marked though late decrease in the sutures. These 
lose the characteristic prominence of the second lateral saddle, which is a pro- 
gressive characteristic in this genus. All the lobes and saddles also become 
broader and decrease in proportionate length, and finally in extreme age the 
abdominal lobe is decidedly shortened? In the Museum of Comparative Zotlogy 
there was also a much smaller specimen,’ in which the same stage of decline had 
been reached at an earlier age. The law of succession was, therefore, quite differ- 
ent from that which governed the inheritance of progressive forms. The most 
retrogressive of the bucklandian varieties were those which were most closely 
connected in every way with Cor. orbiculatum. The genetic connections also 
between Cor. rotiforme and Cor. lyra were traceable only through those varieties of 
rotiforme which had the most convergent sides and the most retrogressive pile, 
tubercles, ete. This also holds for the connections between this last and Cor. 
Gimuendense and Cor. trigonatum. 
The law of succession in eatagenesis, therefore, ts that retrogresswve species in each 
separate genetic serves are the direct descendants of retrogressive varieties or forms. The 
facts consequently are in strict accord with the theory of descent with modification. The law 
+P) ve fig. 8,95 pl vietig. 8; ply vit. fig, 1. ci iAp ih siten ae SOP ew odige 1, 2. 
Sg 
