408 
three factors, viz., rapid spreading of the whorl, the abrupt curva- 
ture and contact or close proximity of the paranepionic stage to the 
apical part of the conch. Even, however, if this conclusion be 
doubted and if, in a few forms of extremely specialized nautilian 
shells of these early periods of geologic history, it can be asserted 
that the impressed zone has really become inheritable ; the position 
assumed in this paper, that the impressed zone is mechanically gen- 
erated in the later stages of growth and becomes an inheritable 
characteristic only in forms with accelerated development, is posi- 
tively strengthened. The whole argument being based upon mor- 
phology, it makes no essential difference how early the impressed 
zone appears or in what form it appears, provided the shells in 
which it is characteristic of the first volution before contact are the 
descendants of those in which this character is transient and 
obviously due to the moulding during growth of one volution over 
the next inner volution. 
My experience, however, in writing this paper has led me to dis- 
tinguish two kinds of impressed zones; that which occurs on the 
free dorsal sides of the young and that which occurs as the direct 
result of contact. I propose therefore to call the former the dorsal 
Jurrow and the latter the contact furrow. 
The ananeanic substage among Carboniferous cephalopods is not 
only marked by the beginning of the contact furrow but also, as a 
rule, by the introduction of correlative changes in the form of the 
whorl. Thus the tetragonal whorl, with an outline similar to that 
of an inverted trapezoid in section, and consequently an obvious 
repetition of the ephebic whorl of Temnocheilus, and with sutures 
also like those of the adults of that genus, appears at this stage in 
Carboniferous cephalopods of several different genera, showing their 
immediate descent from Devonian Temnocheili. 
The first appearance of the dorsal lobe in the sutures is correlated 
with closer coiling and is apt to make its first appearance in primi- 
tive nautilian shells at this stage in the contact furrow. This lobe 
however, occurs also before the whorls touch in a number of forms, 
notably Barrandeoceras of the Silurian, and in one of these, Bar- 
randeoceras Sternbergt, it occurs in the ephebic stage, although 
this is a gyroceran form and no contact furrow is formed. ‘There 
is also another smaller lobe which appears in the centre of this, 
the annular lobe. These are not strictly correlative with the 
impressed zone, since a dorsal lobe appears in some cyrtoceran 
