THE BRACHIOPOD SHELL. 307 



ARTICULATING PROCESSES. 



Among the Atremata, Bicia of the Lower Cambrian shows strong projections on the anterior 

 portion of the cardinal area of the ventral valve beside the delthjrrium (PI. L, figs, li-k and 

 2a), but no corresponding sockets or means of articulation in the dorsal valve. It is a matter 

 of interest to note that bosses occur on both valves (PI. L, figs. 2, 2a-e) close to the front mar- 

 gin of the cardinal area, wliich were evidently of service in connection mth the movement of 

 the valves. In Dicellomus the area of the dorsal valve is sometimes grooved in such a manner 

 (PI. LII, fig. li) as to suggest an articulation ^vith a projection on the ventral valve, and the 

 crenulated margin of figure Ij is suggestive of another tendency toward developing articulation 

 of the valves. 



The actual presence or absence of articulating processes in Kutorgina cingulata (PL V) is 

 very difficult to demonstrate by observation, owing to the character of the matrix in which the 

 specimens occur. That some form of articulating processes is present, however, is indicated 

 by the facts (1) that in a relatively small collection tliirty of the valves are united, and (2) that 

 in only a few instances is the dorsal valve in any other than the normal position in which it 

 would have been held by teeth in the ventral valve. In only a few of the specimens have the 

 valves slid or turned either way, as they usually do in the inarticulates. 



In the Neotremata, the Lower Cambrian Oiolella (PI. LV, figs, le-g) has projections suggest- 

 ing an articulating process on either side of the delthyrium. The same is true of Trematoholus 

 (Pis. LXXXIII and LXXXIV). 



In the Protremata the articulating processes are well developed in Billingsella (PL LXXXV, 

 figs. In, Iv, and Iw), and, so far as known, in Nisusia and the other genera of the order. 



SPONDYLIUM. 



I am inclined to agree with Doctor Schuchert that the spondylium originated as the result 

 of deposition, within the ventral rostral cavity, of testaceous matter about the bases of the 

 adductor, diductor, and pedicle muscles [Schuchert, 1897, pp. 100-102]. With this conception 

 in mind we may consider that the slightly demarked rostral muscular area of Nisusia is one of 

 the earhest known traces of pseudospondylia, and that it is followed by the more decidedly 

 elevated muscular area or pseudospondylium of Billingsella, which is of the same type as that of 

 Clitambonites and essentially of the same general type as that of Eoorthis. In Filkelnburgia the 

 pseudospondylium appears to have been supported at its anterior margin by three septa (PL 

 XCIII, fig. 2), thus forming a link between the pseudospondyhum of Billingsella and the spon- 

 dylium of Syntrophia. In Huenella (PL CHI, figs. Ih, li, 2 1, and 2m) the spondylium is free 

 at the sides and without a supporting septum (as in SyntropTiia, PL CII, fig. 6g; PL CHI, 

 figs. 4d and 4e) or septa (as in ClarlceMa, PL CIV, figs. 2c and 2d). 



Protorthis has no well-observed cardinal process in the dorsal valve, although it has gained 

 a true spondyhum in the ventral; and Syntrophia gains a cruralium in the dorsal valve, both 

 valves of Syntrophia rotundata having either a spondylium or crurahum supported on a median 

 septum (PL CHI, figs. 4d and 4e). In ClarlceTla the spondylium is supported by three or more 

 septa (PL CIV, figs. 2c and 2d). 



In chronologic order the pseudospondylium of the ventral valve first appears in the Lower 

 Cambrian Billingsella highlandensis (PL LXXXVII, fig. 4b) and B. orientalis (PL LXXXVI, 

 fig. 2). It is present in all species of Billingsella from the Middle and Upper Cambrian, and 

 has the same form in the Middle Cambrian as in species of Eoorthis of the Middle and Upper 

 Cambrian. On this line of descent the pseudospondylium appears in Orthis (see Hall and 

 Clarke [1892c, pp. 186-194]) of the Ordovician and later faunas, probably as a reversion from 

 a free spondyhum. On the line of descent to Protorthis the pseudospondylium becomes a free 

 spondylium and continues on through Syntrophia and Clarhella into the Ordovician and Silurian 

 Pentameridse and Clitambonitidse. 



It is not improbable that all Cambrian and Lower Ordovician Protremata have a pseudo- 

 spondylium or spondyhum. In the Orthidse a pseudospondylium is often suggested, but the 



