310 CAMBKIAN BRACHIOPODA. 



the Upper Cambrian affords fine illustrations of the general thickening of the shell beneath 

 the visceral area in both valves, wliile the points of attachment of the muscles may be depressed 

 (PI. VII, figs. 10 and 17) or elevated (PL VII, figs. 9 and 14). The thickening of the shell is 

 also well shown by the ventral valves of Oholus {Schmidtia) (PI. XIV, figs, la, 3a, and 4a) 

 and less well hj the dorsal valves (PI. XIV, figs. Ic, 4c, and 4d). Oholus cijane (PI. XXVII, 

 fig. 4b) from the Lower Ordovician has an unusually well-developed thickening beneath the 

 visceral area of the dorsal valve. Many other illustrations might be given from Oholus and its 

 subgenera. 



In Lingulella and its subgenera there is a striking development of the primitive or 

 early stages of the platforms, and they are present in many species — L. acutangula (PL XVII) 

 and L. granvillensis (PL XXII, figs, lb and Id) of the Lower Cambrian and L. duhia (PL 

 XXIV, 'figs. 4 and 4a) and L. ino (PL XXVI, fig. 4) of the iliddle Cambrian. Of Upper 

 Cambrian species only the small L. ferruginea shows a slight thickening of the valves beneath 

 the visceral area (PL XXIX, figs. Ir and Is). 



From Lingulella there appears to have been an offshoot in Middle Cambrian time in Oholus 

 (Fordinia) perfectus (PL LXIII, figs. 10, lOa-d) that developed a strong tendency toward the 

 formation of a rudimentarj- posterior thickening in the interior of the valves. Two species, 

 Oholus (Fordinia) bellulus (PL LI, figs. 3, 3a-f) and 0. (F.) gilberti (PL LI, figs. 5, 5a-d) 

 from the Upper Cambrian, also show the same tendency. The rudimentary platform of 0. 

 (Fordinia) ■perfectus (PL LXIII, figs. 10a and 10c) is intermediate in development between 

 those of Oholus apollinis (PL VII) and Elkania desiderata (PL LI). The position of the central 

 group of muscle scars, at the anterior sloping ends of elevated ridges, marldng the path of advance 

 of the scars with the growth of the shell, is beautifully shown in figure 10a of the ventral valve 

 of 0. (F.) perfectus (PL LXIII). In the dorsal valve of the latter species there is marked 

 thickening of the shell beneath the visceral area, and a marked median septum or ridge just in 

 front of the cardinal area (PL LXIII, fig. 10c) is very suggestive of a cardinal process. 

 The ventral valve of Elkania desiderata (PL LI, fig. 1) has a strong, solid, rudimentary platform 

 into which the interior cardinal area is blended, and the position of the central group of muscle 

 scars is on the front slope of the platform; in the dorsal valve (PL LI, figs. Ic and Id) a rudi- 

 mentary platform is also clearly indicated. The dorsal valve of Elkania amhigua (PL LI, fig. 

 2c) may have a fairly strong rudimentary platform or be without it (PL LI, fig. 2b). Elkania 

 ida (PL LI, figs. 4b and 4c) has the rudimentary platform finely developed in both valves. 



Lingulasma scJiucherti [Hall and Clarke, lS92c, PL II, figs. 17-23] of the Upper Ordo- 

 vician appears to be the best expression of an intermediate form between Oholus (Fordinia) 

 (PL LXIII, figs. 10, lOa-d) and the forms represented by Dinoholus [Hall and Clarke, 1892c, 

 PL IVb] and Trimerella [Hall and Clarke, lS92c, PL IVa] of the Silurian, but it is quite prob- 

 able that Lingulasma was not in a,nj way intermediate between Fordinia and the trimerellids. 

 It seems to represent a side branch of Lingula. 



The line of development of the platform to Ordovician and Silurian forms seems to have 



been from some Obolus-like form, as Fordinia to Neoholus (PL LXXXI, figs. 2e and 2f), and 



thence to Dinoholus. I thought at one time that Elkania might be a possible ancestral line 



to the Trimerellidje, but, after carefully considering the position of the rudimentary i^latforms, 



I concluded with Schuchert that Elkania was an offshoot from Oholus that did not leave any 



known descendants. 



SEPTA. 



The septum of the dorsal valve of the Obohdse is usually in the form of a narrow ridge, and 



is not a true septum in the usual sense, which is that a septum is a thin median vertical plate 



of variable height, terminating freely or supporting a spondylium. WeU-developed septa occur 



in the Syntrophiidse as supports beneath the spondyUa (PL CIII, figs. 3d and 3e; PL CIV, 



figs. 2c and 2d). 



DENTAL PLATES. 



Dental plates first appear, so far as known, in the Protremata as oDscure processes supporting 

 the teeth of the ventral valve, one on each side of the delthyrium. One of the oldest examples 



