74 Dr. J. Allan Thomson — Brachiopod Morphology. 



uniplication. In other words, ISfeothyris and Hagellcmia belong to the 

 same series of folding, viz. from non-plicate to ventrally biplicate, 

 and on this ground, therefore, there is hardly reason for Neothyris, 

 which, as a matter of fact, has been suppressed by most modern 

 authors. Digressing from consideration of types of folding, I may 

 point out that there are other grounds on which Neothyris may stand 

 as a very good genus. N. lenticiilaris differs from M. flavescens in 

 exactly the same way as Pachymagas tehuelca ' differs from Terebratella 

 dorsata, viz. in the possession of a bifurcating septum and extremely 

 thickened socket-ridges and cardinal process, as against a simple 

 septum, thin socket-ridges, a lamellar hinge-plate, and a small 

 transverse cardinal process. 



Every Recent and Tertiary species referable to Bouchardia, Magasella, 

 Pachymagas, Terebratella, Neothyris, and Magellania belongs to the 

 same series of folding, viz. from non-plicate to ventrally biplicate, 

 and this may be taken as a character of the Magellaninse. 2 Alternate 

 multicostation arising directly from a smooth stage is found on some 

 species of Terebratella and Magellania, but this is superposed on the 

 ventral plication. Schuchert, however, includes in this sub-family 

 the genera Ccenothyris and Ismenia, the former of which is distinctly 

 dorsally uniplicate, while the latter possesses an alternate multi- 

 carination, which, as Buckman has shown, is probably derived from 

 a dorsally uniplicate form. These two genera, therefore, should be 

 removed from the Magellaninse. 



Sub-family Dallinin^. 



' Turning next to the Dallininse, we find a much greater diversity. 

 In Dallina itself, D. septigera (the genotype) is dorsally biplicate, 

 D. raphcelis is the same, while D. floridana is dorsally uniplicate, all 

 belonging to the series lenticular to dorsally biplicate. Terebratula 

 grayi, Davidson, which was also referred by Beecher to Dallina, is 

 a ventrally uniplicate shell with alternate multicostation, and cannot 

 therefore belong to Dallina. There seems to be no difficulty in 

 referring it to Magellania, as it apparently possesses the Magellaniform 

 hinge-plate and cardinal process. Now Dallina and its forerunners 

 must all belong to the series non-plicate to dorsally biplicate, but for 

 the immediate forerunner, Terebratalia, Beecher 3 has chosen for 

 genotype Terebratula transversa, Sowerby, a shell which is undoubtedly 

 venti'ally uniplicate. This was the more unfortunate, as the actual 

 species on which he established a different ontogenetic series from 

 that of Terebratella, s.str., was 'Terebratella obsoleta, Dall, which is 

 a dorsally uniplicate form. Terebratula transversa, Sow., and Tere- 

 bratula eoreanica, Adams & Reeve, two ventrally uniplicate forms 

 from the North Pacific which have been referred to Terebratalia, are, 



1 H. von Ihering, Ann. Mus. Nac. Buenos Aires, torn, ii, p. 332, 1903. 



2 Apparent exceptions are the New Zealand Tertiary forms Terebratula 

 gaulteri, Morris, and Terebratella sinuata, Hutton, which are dorsally uniplicate, 

 but these have proved to be smooth Ehynchonellids. 



3 C. E. Beecher, " Bevision of the Families of the loop-bearing Brachiopoda. 

 The Development of Terebratalia obsoleta, Dall": Trans. Conn. Acad., 

 vol. ix, pp. 376-99, pis. i and ii, 1893. 



