300 CAMBKIAN BRACHIOPODA. 



The general resemblance of the Cambrian eoorthoids to certain Ordovician Protremata is 

 so striking and the lines of descent so suggestive that particular attention was devoted to this 

 o-roup. Examination, however, brought out the fact that this apparent relationship disappears 

 when the shell structure of the two groups is compared. 



Sections of the shells of members of the Billingsellidse, of which figure 1 is typical, all show 

 a lamellar structure with indications of more or less numerous, scattered, very minute pores or 

 tubules passing without interruption through one lamella. In some sections the spots mdicating 

 the tubules are arranged in rows radiating from the beak of the shell to the margins, but no other 

 regular arrangement can be seen. The great mass of the shell is made up of a compact, finely 

 granular base with dark spots and occasional minute crystals of calcite, a groundmass which, 

 under the microscope, appears very much like that of a fine argillaceous shale. 



The Ordovician Protremata have a clearer, more crj'stalline aspect or color than the Cam- 

 brian Billingsellidse, a difference which probably indicates either a purer lime composition for 

 the former or more probably a higher percentage of calcium phosphate for the latter. In 

 chemical aspect the shells of the Billingsellidse appear to resemble those of the Atremata and 

 Neotremata more closely than do the Orthidas. (Compare figs. 1, 3, and 4 with 2.) 



Anatyses of the respective shells would be necessary to prove these relations, but they are 

 interesting subjects for investigation in view of the possible derivation of the Billingsellidge from 

 the Atremata. 



In the Cambrian articulate genera, with the possible exception of Syntrophia and Huenella, 

 the mmute, fibrous structure so characteristic of most if not all orthoids is entirely absent. 

 (Compare figs. 1 and 2.) SyntropTiia and Huenella, however, greatly resemble each other. 

 Thus sections of the shell of Huenella ainormis of the Upper Cambrian (fig. 13) and SyntropMa 

 lateralis of the I^ower Ordovician (fig. 11) show the same radial arrangement of the pores seen 

 in the Billingsellidse, but the shell structure is fibrous and the rows are coincident in direction 

 with the fibers. On closer study this apparently fibrous structure can be resolved into more or 

 less parallel bands or walls of shell substance separating rows of closely arranged, rectangular, 

 porelike spaces. These spaces may be seen distinctly in thick sections, but they disappear when 

 the section is made sufficiently thin to give a clear image under very high power. Sections of 

 the linguloid genera were also prepared and studied, but the thinness of the shells and their 

 phosphatic character prevented very satisfactory results. The irregular large tubules men- 

 tioned by Mickwitz [1896, p. 106] are beautifully shown in sections of Oholus apollinis before me. 



Some of the tubules penetrate several lamellse of the shell and suggest the tubules of some 

 of the orthoids. (See figs. 15 and 15a.) The same general structure, with the exception of 

 the larger tubules, appears to be characteristic of all of the corneous shells of the Atremata and 

 Neotremata, and, so far as knoA\m to me, of all of the Cambrian corneous shells. 



Figures 5-14 are from photographs which have not been retouched, so that personal equa- 

 tion maj^ be eliminated. Unfortunate^, higher magnifications could not be used without a loss of 

 clearness, but even at the present magnification these views show decided differences in structure. 



In conclusion, it appears that the Cambrian Billingsellidse are further removed from the 

 Ordovician and later Protremata than has hitherto been suspected, the microscopic sheh struc- 

 ture in the former being of granular material pierced by small pores and in the latter of fibrous 

 material. On the other hand, the microscopic structure of the Cambrian and later Pentameracea 

 is so similar that an unbroken line of descent is indicated. 



SURFACE ORNAMENTATION. 



GENERAL STATEMENT. 



The known protegulum of all brachiopods is smooth, and the shells have usually advanced 

 well into the nepionic stage of growth before the outer surface is ornamented to any considerable 

 degree. Many species of Cambrian inarticulate brachiopods (especially the Obolacea) remain 

 smooth through all stages of growth, with the exception of lines and varices of growth. Many of 

 them, however, have some form of ornamentation, and some of the most ornate types of surface, 



