308 CAMBRIAN BRACHIOPODA. 



dental plates do not appear to join to form an attached (pseudo) or free spondyUum. The 

 points of attachment of the adductor (central) muscle scars appear to have been directly on 

 the surface of the valve and not on a raised platform or pseudospondylium. 



At my request, and with all the illustrations of this monograph for study, E. O. Ulrich, 

 who has been studying the Ordovician Protremata, prepared the following notes on the spondylia 

 of the Cambrian and Ordovician Protremata: 



SpondyUum. This structure, if we regard the term spondylium as referring only to the typical free or medially 

 supported umbonal camera or spoon, is of more common occm-rence among articulate brachiopods than is usually 

 believed. Hall and Clarke [1892a, pp. 328-335] discuss the theoretic relations of the spondylium to the deltidium 

 and suggest an expansion of the former term that seems to me to be disproved by the chronogenesis of the articulate 

 brachiopods. 



The typical spondylium is more or less rhomboidal in outline and corresponds to a ventral muscular area which 

 is raised above the floor of the valve and formed by the convergence and union of the dental plates. To the concave 

 surface of this structure the ends of the adductor, diductor, and pedicle muscles are attached. 



The manner in which the spondylium is attached to the bottom of the valve is so variable that the feature does 

 not seem to be of more than generic consequence. It is free in Protorthis and in a related new Ordovician genus; sup- 

 ported by one or three septa in Clitavibonit.es, Pentamerus, Syntrophia (s. s.), Clarlcella, etc. ; partly sessile but anteriorly 

 supported by one or thi-ee septa in Finkelnburgia; and wholly sessile (resting entii'ely on the bottom of the valve) in 

 Otusia and Orusia, Eoorthis, Billingsella, Huenella, and certain Clitambonitidse, Syntrophiidae, and Pentameridse. The 

 muscular area of the ventral valve of the Cambrian Orthis-like brachiopods never agi'ees perfectly with that of any of 

 the described Ordovician Orthidse. In the latter the dental plates of the ventral valve never meet and the area is 

 more or less distinctly bilobed and very rarely elevated, though it may be so depi'essed in thick shells as to suggest a 

 spondylium. In the Cambrian types in question the muscular area is often elevated, commonly suggests and fre- 

 quently forms a true spondylium, and as a rule is more or less distinctly rhomboidal in outline., These features, 

 together with the arrangement and components of the muscle scars of the dorsal valve, suggest a relationship with 

 the Ordovician Clitambonitidse rather than with the true Orthidee.o 



Of the true Ordovician orthids, it seems to me that the groups represented by 0. callactis-tricenaria and Dalmanella 

 (especially the D. suhxquata section) are the oldest and most persistent. Both of these types often retain unresorbed 

 remnants of the deltidium, and in the development of their dental plates and septa and in the shape and arrangement 

 of their ventral muscular areas they agree with the average Cambrian articulate more closely than do any other Ordo- 

 vician or Silurian orthoids, with the possible exception of Plectorthis. Thus, except in the matter of surface plication, 

 the resemblance between Orthis tricenaria and Billingsella romingeri (Barrande) is rather striking. But it is to be 

 noted that the muscular area in the ventral valve of 0. tricenaria (as in Dalmanella suhiequata) is blunt and gently 

 bilobed or obscurely trilobed anteriorly, and not subrhomboidal in outline. Plectorthis whitjieldi has longer and broader 

 ventral adductor scars than any other Ordovician or Silm'ian orthoid known to me, and on this account its muscular 

 area resembles that of Billingsella more closely. Plectorthis whitjieldi is, however, of late Ordovician or Silurian age 

 and lived at a time of very rapid evolution (and possibly reversion) among brachiopods, when other orthoids of the 

 same type (Orthis (Billingsella) laurentina (Billings) and Orthis flabellites) were reverting to ancestral characters; or it 

 may be that they were immigi'ants from some other area where the older characters persisted. 



This difference in shape of the ventral muscular area in the Billingsellidse and Orthidfe is due to the fact that in 

 the former the median (adductor) scars are always longer and as wide or wider than the lateral (diductor) pair. In 

 the Orthidee the adductor scars are of the same general shape as the corresponding scars in the Billingsellida, but are 

 relatively smaller and shorter. They are as long as the diductors in Orthis tricenaria, Plectorthis whitfieldi, and Dal- 

 tnanella subiequata, and in others [Dinorthis, Rhipidomella, etc.] they are much shorter and more or less completely 

 inclosed anteriorly by the revolutionary tendency of the area of the diductors to expand at the expense of the adductors. 



A natiual division of the orthoids into two families may thus be established: 



(1) Ventral muscular area small, obovate or obcordate; adductors reaching front margin of area (Orthis (s. s.), Plec- 

 torthis, Platystrophia, Hebertella, Orthostrophia, Dalmanella). 



(2) Ventral muscular area large, bilobed or elliptical; adductors proportionately small and more or less com- 

 pletely inclosed anteriorly by the flabellate diductors [Heterorthis, Plxsiomys? deflecta group, Dinorthis, Bilobites, 

 Rhipidomella, ScMzophoria, Orthotichia]. 



Another well-marked difference between the true Ordovician orthoids and their billingsellid ancestors is in the 

 arrangement of the muscle scars of the dorsal valve. In the Billingsellidse they are somewhat elongate and more or 

 less radially arranged (PL LXXXVIII, figs. If-k; PL XCIII, fig. If; and PL XCIX, fig. 2d) and in the Orthidas [Orthis 

 tricenaria, Plectorthis, Dalmanella, and Orthis tritonia] they are shorter and placed longitudinally. Some of the later 

 orthoid forms in the second group [Bilobites, Rhipidomella, and Schizophorial are characterized by a radial arrangement 

 of the dorsal muscle scars, but the ventral muscular areas are here highly developed and bear little resemblance to 

 those of the Billingsellidse. 



" Commenting on this, Dr. Charles Schuchert wrote me: "This is a very important point. It has often strucls: me that all the early Cambrian 

 brachiopods have an incipient spondylium, sometimes free, sometimes supported, or sessile. Out of the sessile forms have developed the true 

 orthids." 



