.ToHNSON — is Archceopteris a Pteridosperm ? 125 



escape of tlie contents of the sporangia would be facilitated. The stalk of 

 the sporangium, i.e. the sporaugiophore proper, is vascular. A single bundle 

 may be seen entering it from the sporophyllule, and passing to the base 

 of the sporangium itself (Text, fig. 1). (I have seen as a rarity a possible 



indication that the bundle is continued as a ridge along the sporangium surface 

 towards its apex.) In nearly all living Filicinese, as well as in other Pteri- 

 dophyta, the sporangium, and its stalk, when present, are non-vascular. The 

 Ophioglossacese, and less markedly the Osmundacese, are striking exceptions, 

 it may be noted in passing. 



It seems desirable to use, as I have done, the convenient term sporangio- 

 phore in a non-morphological sense, simply to indicate the stalk, whether 

 vascular or not, carrying tlie sporangium or synangium, and to leave 

 morphological interpretations to be expressed in each group by otiier special 

 terms. Thus in the case of Archseopteris the fertile pinnule or sporophyllule 

 is clearly the homologue of the sterile pinnule. Tlie oblong-oval, fusiform 

 or club-shaped esannulate sporangium sometimes presents a surface which is 

 uniform, except for a fine longitudinal stiiation due to the walls of the cells 

 forming its outer layer. Very rarely there is a distinct more or less median 

 groove, suggestive of a longitudinal slit and observable as a slit or crack in the 

 carbonaceous impression. More frequently there is a furrow running along 

 the two sides of the sporangium. In many cases the central carbonized 

 matter of tlie body of the sporangium has disappeared, and is represented in 

 the rock by a vertical row of 3-6 distinct pockets, separated by transverse bars 

 (PI. v., fig. 4, PL VI.) . The carbonized body of the sporangium is seen divided 

 transversely into five or six sections by horizontal grooves, which can scarcely 

 be due to artificial splitting caused by shrinkage of the carbonaceous crust. 

 Sometimes the carbonaceous wall shows cross-bars apparently, at similar 

 intervals. Occasionally the surface is tuberculate, as if raised by the contents 

 of the sporangium. As these, in spite of various attempts, have not y^t been 

 seen it would be pure conjecture to suggest that sporangia with a warty 

 surface may have contained large spores and the less irregular ones small 

 spores. There is, however, clear evidence that the sporangium is not a 

 simple spore-capsule opening by an apical pore as suggested by one writer. 

 It is, it seems, divided by transverse septa into a multilocular spore-capsule, 



