scHucHEET.] DEVELOPMENT OF THE PROTREMATA. 81 



from the Falls of the Ohio, These are believed to be actual and not 

 chance attachments. In Crania cementation occurs very early and is 

 complete, causing all obliteration of the iirotegulum and subsequent 

 stages of growth in the ventral shell. That cementation does obliterate 

 nearly all the younger characters is also shown in the remarkable gen- 

 era Richthofenia and Ostrea. On the interior of Pholidops and Crania 

 the four large muscular scars, which are more those of the ISTeotremata 

 than of the Atremata, are arranged medially, in the center of which, 

 probably, was the pedicle opening. Some proof of this is seen in the ex- 

 cavated, posteriorly terminating muscular pit of Crania {(/nabergensis, 

 which, if carried through the valve, will make the pedicle opening 

 snbcentral and surrounded by shell deposit. If an Acrotreta, Linnars- 

 sonia, or Conotreta became cemented, there would result practically a 

 Crania, In no atrematous brachiopod is there the slightest indication 

 of cementation, but where shell fixation does occur it is always (ex- 

 cepting in Zugmeyeria and Thecocyrtella) in such as have the pedicle 

 very early surrounded by shell matter, as in the Stroj)homenid;t and 

 Productidie, For these reasons the characters of Craniacea seem 

 more in accord with the Neotremata than with the Atremata. The 

 characters of Craniacea are certainly not of ordinal importance, and 

 possibly not even of superfamily value. 



In the development of its pedicle foramen the family Siphonotretidae 

 is unlike any other of this order. During neanic growth the pedicle 

 opening was posterior to the protegulum, but later it gradually moves 

 anteriorly through the shell by resorption, producing a narrow slit 

 similar in appearance to that of the Discinidie, A pedicle foramen of 

 the same nature is also developed in Eichwaldia and Dictyonella of 

 the Protremata. As yet no explanation has been given as to the 

 causes producing this aberrant development. The writer suggests 

 that since these animals had delicate peduncles, with the shell elon- 

 gate oval and sometimes cone-shaped in form, they probably stood 

 nearly upright on their pedicles in early growth. Shell accretion 

 being more rapid anteriorly, with the ventral side of the animal the 

 larger and heavier, a tendency Avas initiated for the shell to lean 

 against the ventral side of the j)eduncle. This pressure would ijroduce 

 resori)tion of the ventral shell anterior to the pedicle, and eventually, 

 this tendency becoming hereditary, the ventral valve would lie nearly 

 flat, with the pedicle emerging at*a great angle subcentrally. 



PR0TRE3IATA. 



This order is represented by 738 species, or nearly 40 per cent of 

 American Paleozoic brachiopods, and is eminently characteristic of the 

 post-Cambrian Paleozoic systems. Like the Atremata and Neotre- 

 mata, it is represented in the Lower Cambrian. It was not, however, 

 until Ordovician times that the Protremata attained very rapid evolu- 

 tion. In the Cambrian there are but -i genera and 22 species, while in 

 Bull. 87 6 



