1887] Geology and Paleontology. 65 
ran, and finally to a close-coiled nautilian shell. Among Ammo- 
noids the same series occurs only on one occasion, at the begin- 
ning of the group, during Silurian and Devonian time, in a series 
which may be said to include Bactrites, a straight orthoceratitic 
shell, Mimoceras, a true gyroceran form, and Anarcestes, which 
is close-coiled. The discovery of a proto-conch upon the apex 
of Bactrites by Beyrich and Branco leaves no doubt that it is, as 
heretofore supposed by the writer, a transitional form from Or- 
thoceras to Ammonoidea. These forms are primitive or transi- 
tional radicals and have cylindrical whorls, except in Anarcestes. 
In this genus a depressed semilunar whorl is for the first time 
introduced. This form of whorl is not at once and generally 
adopted in the young. On the contrary, these are usually 
tubular and often straight like Bactrites, or loosely coiled like 
the adults of Mimoceras. Others, again, after passing through 
a stage with tubular whorls, may become suddenly close-coiled 
and have at once a depressed form of whorl. Such fluctuations 
in embryonic characters are common even in different varieties 
of the same species until we reach the Trias. In this formation, 
or possibly earlier in the Dyas, the larve are all close-coiled, and 
the whorls at an early stage invariably have the depressed semi- 
lunar form like the adults of Anarcestes. Throughout the Trias 
also there occur in great abundance smooth shells, Arcestes, in 
which the full-grown adults are smooth and have the similar 
anarcestian peculiarities. Thus from the Silurian to the Trias, 
inclusive, the semilunar or depressed smooth whorled forms are 
which we have designated as primary radicals, confining the use of 
the words primitive radicals to the transitional genera Bactrites, 
Mimoceras, and the like. 
Compressed forms differing but slightly from the depressed 
species occur in Anarcestes and in Arcestes, etc. In the Trias 
- and Lias these compressed, smooth shells which we have called 
secondary radicals become much more important. In Psiloceras 
planorbe we strike upon a species of this character to which we 
can trace all the Arietidz of the lower Lias and many forms of 
higher Jura and Cretaceous. 
The great trunk of radical species has, of course, many lateral 
branches, which strike off from it during the course of its chrono- 
logical migrations through the Palzozoic and Trias, but of these 
we have taken no account, because they were purely lateral off- 
shoots which did not arise from fission or the modification of 
the main stock of radical generators. In the Jura, however, this 
main stock itself splits into branches, and the pri and — 
secondary radical forms are replaced by more complicated radi- 
cals. 
There is a side branch, which arose in the early Trias, aid i in 
which they are still, in a measure, preserved and continued, but 
VOL. XXI.—NO. I. 5 
` 
