scHucHEET.] THE SPONDYLIUM. 101 



The writer also previously entertained this view, but when it became 

 known that s])on(lylia are developed where no dental lanielhe exist, as 

 in the Lingulasmatidte and Trimerellidte of the Atremata; that spon- 

 dylia are never present in the Neotremata, where a pedicle-sheath 

 is sometimes well developed, as in the Acrotretidie; and finally, that 

 a si)ondylium is even present where no deltidium ever existed, as in 

 the two first-mentioned families, and in Oyrtina, Oamerospira, Merista, 

 and Dicamara of the Telotreinata, such an explanation became unten- 

 able. The fact that solid or excavated spondylia exist in three orders, 

 two of which never developed a pedicle-sheath (Atremata and Telotre- 

 mata), and one had no dental lamelhe (Atremata), is good evidence that 

 the jirodeltidium i)rimarily had nothing to do with the development of 

 spondylia. Further, no spondylia are developed in the Cambrian until 

 long after the deltidium was well established, and therefore the spon- 

 dylium can not be "but a modification of the original xjedicle-sheath." 

 However, it is very probable that when the dental lamelhe in the Pro- 

 tremata became sufficiently wide to Join the ventral shell, crowding all 

 the muscles of this valve into a small area, these took advantage of the 

 inner sides of the dental lamelhe for insertion, and thus a continuous 

 layer of testaceous matter was deposited within the rostral cavity. 

 With growth, the muscles move forward and press against the genitalia, 

 which causes resori^tion or nondeposition for their relief. Xo spondylia 

 appear before the Upper (/'ambrian, and here also are the first com- 

 pletely developed dental lamellte. The so-called Lower Cambrian 

 camarellas have no comi>letely developed dental lamelhe, and are 

 related to the rhynchonelloid genus Protorhyncha, and to Protorthis 

 billiuffsi, which also has no spondylinm.' Therefore, the further conclu- 

 sion of Hall and Clarke can not be accepted, that, "where the teeth 

 are wholly without dental lamelhe, or where such lamella' do not extend 

 to the bottom of the valve, it seems necessary to regard them as 

 instances of degeneracy or resorption of the primitive spondylinm."^ 



It seems clear to the writer that since the "shoe-lifter" j)late, or 

 spondylinm, in Merista and Dicamara is for nuiscular insertion, this 

 plate in the ventral valve of these genera is the morphic equivalent of 

 the spondylinm in the Pentameracea, and that the dorsal muscular 

 plate in Dicamara is the equivalent of the cruralinm, and can not "be 

 interpreted as an entirely different structure from the spondylinm."-' 

 It is true that the siwndylia of these genera are not exactly like those 

 of the Pentameracea, but since this plate in the Atremata is not formed 

 by the union of dental lamella', as these do not exist in this order, there 

 is no reason for rejecting the terminology for these plates in Merista 

 and Dicamara. 



> Oamarella minor and O. antiqua are more closely related to Protorhynclia than to any other genus. 

 Of Orthis billingsi, the typo of Protorthis, very good casts of specimens in the Cornell University 

 Museum are in the National Musinim, which show that this genus also has no spondylium, and that its 

 characters are those of Billingselhi. 



= Hall and Clarke, ihid., p. 333. 



sibid,, p.335. 



