12 OLAF HOLTEDAHL. [SEC. ARCT. EXP. FRAM 



land, observed quite prominent spines in this older specimen too. This 

 Gotland-specimen differed from the Arctic specimen chiefly in having less 

 numerous tabulae, the distance between two of them generally equalling 

 the width of the cell. The closeness of the tabulae is very conspicuous 

 in the Ellesmereland form as seen in pi. IV, fig. 6; the type specimen of 

 NICHOLSON may also have as many as 7 tabulae in the space of two lines. 



Two forms which probably may be regarded as identical with the 

 one here described are the Favosites pyriforme HALL as described 

 and depicted by STUART WELLER 1 from the Decker Ferry Group of New 

 Jersey and by SwARTZ 2 from the Keyser member of the Lower Helder- 

 berg of Maryland. 



The original Astrocerium pyriforme HALL (Pal. New York. Vol. 2, 

 1852, p. 123, pi. 34 a, figs, la 1 e) as early as in 1899 is placed by 

 L.AMBE 3 under Favosites hisingeri E. and H. 



HALL'S form with its very long spines seems also to be decidedly 

 different from the two forms just mentioned, and from the Arctic 

 species. 



Occurrence: B, upper part, point north of the Beautiful Valley, 

 Walrus Fjord. 



Favosites sp. 



A fragment of the. uppermost highly convex portion of a very 

 regularly built Favosites, that differs from the two preceding species, 

 is visible in a small piece of rock in the material. The nicely radiating 

 corallites have mostly a hexagonal section with an average diameter of 

 about 2 mm. Rarely, we observe smaller ones between. The tabulae 

 are not so closely arranged as in the previously mentioned species. Only 

 one or less occurs per mm. The pores in this specimen are discernable. 

 They are comparatively large, show an elevated outer rim, and are dis- 

 posed in two rows. 



On account of the fragmentary condition no specific determination 

 can be made with certainty. With the tendency to variation of the specific 

 characteristics of these Favosites-lorms it is evident that it is valueless 

 to make detailed identifications on very scanty material. 



Occurrence: B, upper part, over the fragment-limestone, valley 

 south of Borgen. 



1 Geol. Survey of New Jersey, vol. 13, 1903, p. 220, pi. 17, figs. 3-5. 



2 Geol. Survey of Maryland. Lower Devonian. 1913, p. 211, pi. 23, figs. 1 4. 



3 Contributions to Canadian Paleontology. Vol. IV, Part I, 1899, p. 6. 



