26 T. G. HALLE, LOWER DEWONIAN PLANTS FROM RÖRAGEN IN NORWAY. 



mate branches are sometimes very short so that the sporangia appear to be sessile. Daw- 

 son (1. c. ) speaks of the sporangia as »either lateral or terminal »; but each capsule is prob- 

 ably always placed in the top of a branch, though this may be very short." The best 

 preserved specimen consists of a bifnrcating lateral branch, attached at almost a right 

 angle to a much broader branch or axis which is broken just above the point of branch- 

 ing. One of the sporangia appears to show a longitudinal split, possibly indicating 

 the mode of dehiscence, but it is not certain that this is a natural feature. - - In the spe- 

 cimen in fig. 8 (thrice enlarged), which is from the same piece of rock, the sporangia 

 appear unusually short and rounded at the apex. They are not well preserved, ho we ver, 

 and may possibly not be quite developed. — The specimen shown in figs. 3 and 4 (twice 

 natural size) of pl. 3 agrees fairly well with those now described, with the exception that 

 even the thicker branch divides dichotomously. The best preserved sporangium has a 

 dark median longitudinal line, which possibly denotes the line of dehiscence. — The photo- 

 graphs in fig. 5 (natural size) and fig. 6 (thrice enlarged) represent a curved branch bear- 

 ing at least three sporangia, which are so shortly stalked that they appear sessile. 



There is fairly great difference between the specimens now described and that 

 shown, seven times magnified, in pl. 3, fig. 9. It consists of a long curved branch termi- 

 nating in a single sporangium which is nearly lanceolate on t lie impression and tapers 

 gradually to a sharp point. It is possible that this specimen may not be specifically iden- 

 tical with the others; but the material is too scanty to permit of any decision. In other 

 specimens, too, the apex of the sporangium may sometimes be fairly acute (cfr. fig. 6), 

 but not so much as in this specimen. 



In regard to the manner of branching, the two counterparts drawn in pl. 4, figs. 19 

 (natural size) and 20 (twice enlarged), respectively, present a peculiar habit. They 

 show a very slender, repeatedly bifurcating branch-system; and in fig. 20 two long and 

 pointed sporangia are seen, borne in the top of a long, curved stalk. These sporangia 

 agree in shape with that in fig. 9, pl. 3, just described. It is even possible that the resem- 

 blance extends further, as what appears to be two sporangia may really be the two 

 halves of a single sporangium split longitudinal ly. 



Morphological nature and relations. 



The specific identity of the Röragen specimens and those described by Dawson 

 cannot be doubted. The latter present the same variations as our material both as re- 

 gards the manner of branching and the shape and size of the sporangia. A main branch 

 or rachis is developed, for instance, in the specimen in Dawson's pl. 10, fig. 118; and 

 this specimen also shows a similar longitudinal striation. The specimen figured by P. 

 Bertrand is probably identical, too. Another somewhat similar supposed fructifica- 

 tion is figured by Bureau (1911, p. 20; pl. 4, figs. 31, 32; 1914, p. 27; pl. 1 bis, fig. 4, 4 A) 

 from the upper Devonian of the Loire district in France under the name of Cephalo- 

 pteris mirabilis Nath., from which species, however, it seems to differ considerably. The 

 similarity of Bureau's specimens to our species is hardly sufficient to prove a specific 

 identity. 



