264 



ROBERT TRACY JACKSON ON ECHINI. 



then the tubercle is imperforate, and in the youngest plates, no primary tubercle exists. Sec- 

 ondary and very small miliary tubercles are thickly distributed along a narrow border outside 

 the scrobicular circle, and dorsally this border is much larger on the aboral aspect of each plate. 

 Primary spines are stout, inflated in the lower third, tapering from there distally; the milled 

 ring is strongly marked, above which the spine is vertically finely striate for a short distance, and 

 above this part densely studded with short blunt knob-like spinules, set without any definite 

 order. Secondary spines 3 to 4 mm. long are situated on secondary tubercles on the borders 

 of the plates; thej^ are short, flattened, spathulate, vertically finely striate. Between the 

 secondary spines are found some miliaries (Plate 11, fig. 4) which are minute and connected 

 with miliary tubercles. These are the only miliary spines yet seen in the Palaeozoic. The 

 peristome is plated with ambulacral, and many rounded strongly imbricate non-ambulacral 

 plates. The lantern is typical of the Palaeozoic as described below. 



Lower Carboniferous, Miatschkowa, Province of Moscow, Museum of Comparative Zoology; 

 British Museum; Berlin Museum flir Naturkunde, two specimens including the one described 



by Tornquist (1896); Strassburg Mu- 

 seum; Munich Museum; Freiburg i. 

 B. Museum. The var. schellwieni of 

 Tornquist, Fusilinenkalk of the Carnic 

 Alps; Miatschkowa and many other 

 localities in Russia (Eichwald). 



A very choice and complete test of 

 this species is in the Munich Museum 

 (text-fig. 239 bis; Plate 10, fig. 10; Plate 

 11, figs. 1, 2). This is the best pre- 

 served specimen known in the genus. 

 On the peristome there are a few am- 

 bulacral and non-ambulacral plates all 

 small and imbricating adorally. In the 

 basicoronal row of the corona, in the 

 interambulacra of areas A, I, G, the 

 plates of columns 1 and 4 are whole and 

 those of columns 3 and 2 are half plates, 

 in which the ventral half has been re- 

 sorbed. On the other hand, in areas C and E, in the basicoronal row, the plates of columns 

 1 and 3 are half plates, of which the ventral half has been resorbed and the plates of colurans 

 4 and 2 are whole. This, it is believed, as in A. ivortheni (Plate 9, figs. 6-8) is associated 

 with the introduction of columns in development, If, as in the figures cited, column 4 originates 



239 b/s 



Text-fig. 239 bis. — Archacocidaris rossica (Buch) . X 4. Ven^ 

 tral border of corona and peristome, same specimen as Plate 1 1 , fig. 1 



