CHANGES IN THE SHAPE OF THE SHELL. 77 



With regard to the passage of the peduncle out of the shell, 

 Beecher has distinguished four different types : 



I. Atremata. The peduncle passes out simply between the 

 posterior edges of the two valves, retaining the direction of the 

 axis of the body. Lingula, Oholus, Paterina. 



II. Neotremata. The peduncle arises from the ventral shell-valve 

 and stands at right angles to the axis of the body. It lies either in 

 an incision of the ventral valve [ScMzocrania), or becomes surrounded 

 by the ventral valve as it grows, so that it then emerges from a 

 subcentral orifice in the valve. Orhiculoidea, Discina, Acrotliele. 

 This group is perhaps nearly related to the Craniidae. 



III. Prototremata. The peduncle arises from a triangular del- 

 tidial slit in front of the l)eak of the large shell, which may either 

 be open (Orfhis, TrojndolejJtus) or closed by a pseudodeltidium 

 (OrfJiisina, Lejjfaena, Stropliomena, Chonetes, Strojiliaeodonta). The 

 Thecidiidae also belong here. 



IV. Telotremata. The peduncle arises from a foramen perfo- 

 rating the beak of the ventral valve. The orifice from which it 

 emerges is, as a rule, surrounded by a paired deltidium {S^nriferidae, 

 Atrypidae, Rliynclionellidae, Stringocephalidae, and TerehratuUdae). 

 In these grou^ps a calcareous brachial skeleton develops. 



With regard to the development of the closing pieces of the 

 beak-orifice known as p/^eudodeltidium and deltidium, li(;tle has been 

 discovered. Beecher, following up Kowalevsky's observations on 

 Thecidium, traces back the pseudodeltidium to a skeletal secretion 

 of the peduncle. During the metamorphosis of the attached larva, 

 the two folds of the mantle are re flexed anteriorly (Fig. 40 J.) and 

 yield the first rudiment of the shell (the protegulum, ds and vs, 

 p. 78), the line at Avhich the reflexion of the valves took place 

 indicates the hinge-line [s) of the protegulum. In Thecidium, the 

 dorsal side of the peduncle-rudiment now also becomes covered 

 with a skeletal secretion {p), Avhich in later stages fuses with the 

 growing beak of the ventral valve (Fig. 40 B and C) and becomes 

 the pseudodeltidium. The true deltidium of the Bhymhonellidae 

 and the TerehratuUdae, on the contrary, is probably to be derived 

 from the reflected parts of the mantle of the ventral shell-valve 

 itself. 



The two shell -valves, when they first appear, are completely 

 separated from one another by the peduncle that emerges between 

 them. They come into contact first at the outermost ends of the 

 straight hinge-line (Fig. 36, p. 73), and at this point the first 



