KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 57. N:0 |. 7 



gives the following diagnose of the genus and the single known species, Arthrostigma 

 gracile Daws.: 



»Stems elongated, cylindrical, bifurcating, and giving off lateral branches; irre- 

 gularly furrowed or ribbed longitudinally, with circular leaf-scars arranged in whorls, 

 and bearing linear rigid leaves with circular bases. Structure apparently cellulär, with 

 a slender vascular axis; fructification probably in cylindrical strobiles. » 



That this diagnose is not quite to be relied on is evident from a comparison with 

 the following passage in the appended description of the material: »The stems were not 

 observed certainly to bifurcate, though there are indications of this; but on one a branch 

 jilaced nearly at right angles was observed. » The supposed »fructification » is in itself 

 of doubtf ul nature ; and there is not the slightest evidence that it belongs to this species, 

 except the uncertain one of association in the same beds. It has therefore not generally 

 ben accepted as throwing any light on the real nature of Arthrostigma. Another import- 

 ant character, and indeed the one from which the name of the genus has been derived, 

 viz., the arrangement of the leaves in whorls, must be subject to modification, as already 

 remarked by C. E. Weiss. This question will be more fully treated below. 



In 1889 a discussion arose between Dawson and C. E. Weiss in the »Zeitschrift 

 der deutschen geologischen Gesellschaft » (pp. 167, 553, 554) regarding the relation be- 

 tween Arthrostigma gracile Daws. and Drepanophycus spinaeformis Göppert (1. c). 

 Drepanophycus spinaeformis is the oldername (1852); but Dawson maintains that even if 

 the two plants were proved to be identical, it could not be used, because it implies a fucoid 

 nature, whereas Arthrostigma is regarded as a land-plant. Weiss (1889, p. 168) suggested 

 the name Drepanophytum to replace Drepanophycus; but this, as Dawson remarks 

 (p. 554), should then date from 1889, while the name Arthrostigma was proposed in 1871. 

 Weiss objects to the name Arthrostigma because the stem is not really articulated with 

 the leaves in whorls as in Dawson's restored figure (1871, pl. 13, fig. 155). The reasons 

 for retaining Arthrostigma in the present paper will be stated below. 



Besides at Gaspé and Campbellton in Canada the species has låter been identified 

 by Kjdston (1. c.) in the Lower Old Red Sandstone of Perthshire in Scotland. The Scott- 

 ish specimens, which differ somewhat from those of Dawson, liad previously been de- 

 scribed by Jack & Etheridge (1. c.) as Psilophyton ? sp. 



In contrast to the amount of discussion which has been going on concerning the 

 true nature of Psilophyton — that other problematic genus of supposed Lower Devonian 

 land-plants — Arthrostigma has received but scant attention from palaeobotanists. This 

 may be partly due to the fact that Arthrostigma so far has been reported only from a 

 couple of localities; and it is therefore fortunate that Prof. Goldschmidt's discovery has 

 put us in possession of more material of this interesting genus. 



Description of specimens. 



A considerable number of the specimens collected in the Röragen area by Prof. 

 Goldschmidt and myself appear to belong to Arthrostigma gracile. The material displays 

 a good deal of variation, and the extreme forms are very different. As, however, a gra- 

 dual transition can be traced between the different types, it is difficult to establish any 



