44 T. G. HALLE, MESOZOIC DEPOSITS AND FLORAS OF PATAGONIA AND TIERRA DEL FUEGO. 
form. The great resemblance in regard to the shape and attachment of the apical 
projection becomes very evident on a comparison of the side-views of the scales 
shown in text-fig. 4 and pl. 3, fig. 15. As far as the structure of the fossil form 
is known there cannot be stated any difference sufficient for generic separation. 
While admitting that the most important characters, such as the number and attach- 
ment of the seeds and the structure of the seed-scale, cannot be made out in the 
fossil form, the latter would appear to be really closely allied to Arthrotaxis and to 
be best referred to under the generic name ÅArthrotaxites. 
It may be stated in this connection that the genus Arthrotaxopsis of FONTAINE 
(1889) has not been shown to be at all closely related to Arthrotaxites UNG., though 
the vegetative shoots present a certain resemblance. 
All the specimens of Arthrotaxites Ungeri are from locality c at Rio Fösiles. 
In this connection may conveniently be described a specimen (pl. 3, fig. 21) 
which appears to represent a male strobilus of a conifer. It is very small, meas- 
uring only 7 mm. in length, with an ovate outline on the impression. It consists of 
a rather slender axis bearing bract-like appendages. These consist of a lower almost 
horizontal petiolar portion and an upper part which is directed almost straight up- 
wards. At the rather sharp bend there appears to be a small dorsal projection, 
direceted downwards. 
It appears at least highly probable that the interpretation of this specimen 
suggested above is the right one. There is no trace to be seen of any pollen-sacs, 
and the structure cannot be made out except in the respects stated; but it must be 
admitted that the fragment resembles nothing so much as a coniferous male strobilus. 
Arthrotaxites Ungeri being, with the exception of the fragment described below, the 
only conifer of the San Martin flora and besides a very common one, represented 
both by vegetative shoots and female cones, it is very probable that this hypotheti- 
cal male strobilus belongs to the same species. A certain confirmation of this prob- 
ability may be recognized in the fact that the specimen shows a considerable habitual 
resemblance to the male strobili of the recent genus ÄArthrotaxis, as exemplified, for 
instance, by the figure of AÅ. cupressoides in ENGLER & PRANTLI'S »Naturliche Pflanzen- 
familien». 
The specimen was found associated with remains of Arthrotaxites Ungeri in 
locality c at Rio Fösiles. 
Elatocladus sp. 
Pl. 5, figs. 7—9. 
The only conifer in the San Martin flora, in addition to the species described 
as ÅArthrotaxites Ungeri, is the little fragment figured in pl. 5, fig. 7. This represents 
only a small portion of the upper part of a branch and measures not quite 3 cm. 
in length. The leaves, though probably spirally arranged, are spread out to two 
