CARPENTER ON TIIE STUDY OF THE EORAMINIEERA. 331 
often happens in the spiral type) are further divided into secondary 
* chamberlets.’ But while, in Orbitolites, these freely open into one 
another, this is not the case with Cycloclypeus , so that its direction of 
growth, as Dr, Carpenter observes, “ is in reality rather radial than 
cyclical.” 
Transitional conditions between these three leading inodes of 
growth are readily found. Thus the straight shell of Nodosaria is 
connected with the spiral Cristellaria by a number of intermediate 
forms so gradually increasing in curvature that the entire series has 
been included in a single comprehensive genus,— Nodosarina. So, 
likewise, in the nautiloid Peneroplis we often notice that “ the spire 
“ becomes disengaged, and prolongs itself in a straight line, its succes- 
“ sive chambers exhibiting little or no increase in size.” Between the 
spiral and cyclical types, on the other hand, the genera Orbiculina 
and Heterostegina may be said to inosculate. Both shells commence 
the multiplication of their chambers after the spiral mode, and “the 
“ earlier convolutions invest those which precede them, whilst the latter 
e: extend themselves peripherally instead of centrally, and the mouth 
“ of the spire or 4 apertural plane ’ widens on each side so as to pass 
“ round (it may be) the whole circumference, and thus to convert the 
“ spiral into the cyclical plan of growth.” These two discoidal shells, 
in spite of their very different aspect, are intimately related to 
Alveolina and Fusulina , respectively, in which the axis of the spiral 
becomes more or less elongated, and the shell, in consequence, assumes 
a fusiform or ellipsoidal figure. 
The spiral plan, on the whole, may certainly be said to predomi¬ 
nate. A majority of the Polythalamia exhibit this mode of growth, 
and this only; and in comparatively few generic forms has it not yet 
been observed. While both the straight and cyclical types, as we 
have seen, are often spiral at some period or other of their life. 
Even Orbitolites is liable occasionally to commence its growth after 
the spiral method. 
Orbitolites is further remarkable for its tendency to pass into a 
complex type which, though connected with the simpler form by un- 
mistakeable gradations (sometimes observable even in different parts 
of the same specimen) offers, nevertheless, in its extreme condition, 
a well-marked modification of the polythalamous structure. Here 
the simple disc is replaced by two ‘ superficial layers ’ and an inter¬ 
mediate stratum, and the margin of the much thickened shell is 
furnished with several series of pores, placed one above the other. 
The chambers of the superficial layers are oblong, not rounded, as 
in the simple type, and their mode of communication with one another 
horizontally is likewise different. Vertically they are brought into 
connection by columnar sub-segments of sarcode, traversing the prin¬ 
cipal thickness of the disc. 
Both the simple and complex types of Orbitolites have obviously 
some relation, however distant, to Dactylopora , which presents us 
with a mode of growth distinct from any of those noticed above, and 
whose structure is, in more than one respect, so peculiar as to render 
