344 PALJEONTOLOGY OF NEW YORK. 



shells, the jugum, however, which is subject to the most frequent variations in 

 form, and which serves as the generic index. When the spirals are directed 

 outward toward the lateral margins of the valves, the jugum seems to be much 

 more variable than in shells where the spirals are introverted or take some 

 intermediate position. In the latter there is a much greater variation in the 

 position of the loop upon the primary lamellaj than occurs in the former. 



The earliest spire-bearing shells yet discovered are the simplest in the 

 structvn-e of the brachidium. Hallina, Protozyga, Cyclospira, of the Lower 

 Silurian, possess brachidia which make a little less than one or two volutions 

 of the calcified lamellaj, with a slight inclination toward each other, and to the 

 median axis of the shell. Zvgospira and Glassia, the contemporaries and suc- 

 cessors of these primitive structures, show progressed conditions of the same 

 form of brachidium. In these genera, however, there is a slight deviation in 

 the vertical axes of the spirals from the transverse axis of the shell, the apices 

 being inclined somewhat toward the brachial valve, and this tendency to lateral 

 evolution in the spiral cones is carried to its extreme in the genus Atrtpa, 

 where the multispiral cones of the fully matured forms of the Devonian may 

 sometimes have their axes nearly parallel. This is the termination of all revo- 

 lution of the cones, a change through an arc of less than 90°, probably due in 

 a large degree to alterations in the form of the internal cavity of the valves ; 

 and the fact that this revolution here ceases, strictly delimits the group of 

 forms bearing spirals to this type {Atuyfwjs). 



It is well to emphasize the fact, lest misconceptions already set on foot should 

 become prevalent, that no wider revolution of the spiral cones exists. It is 

 true that there is a difference of ISO*^ in the position of the axes of the spiral 

 cones in Cvclospira and Spirifer, but the spirals have never, by gradual 

 changes, revolved from their mverted position m the former to their everted 

 position in the latter Such a process might have been possible, but had it 

 actually occurred the forms resulting would have been totally different in 

 structure from any now known. Instead of having the primary lamelliB and 

 jugum on tha dorsal side as in all shells with everted spirals, these parts would 

 lie on the ventral side of the shell. It must hence be inferred that the 



