PAST HISTORIES OF PLANT FAMILIES 1,9 



like those found on the petiole and leaves, which were, 

 indeed, the first clues that led to the discovery of the 

 connection between the seed and the plant Lyginoden- 

 dron. 



The pollen grains seem to have been produced in 

 sacs very like fern sporangia developed on normal 

 foliage leaves, each grain entered the cavity / <; in the 

 seed (see fig. 56), but of the 

 nature of the male cell we are 

 ignorant. In none of the 

 fossils has any embryo been 

 found in the " seeds ", so that 

 presumably they ripened, or 

 at least matured their tissues, 

 before fertilization. 



These, in a few words, 

 are the essentials of the struc- 

 tures of Lyginodendron. But 

 this plant is only one of a 

 group, and at least two other 

 of the Pteridosperms deserve 

 notice, viz. Medullosa, which 

 is more complex, and Heter- 

 angium, which is simpler than 

 the central type. 



HeterangiiiTn is found also 

 in rocks rather older than the 



coal series of England, though of Carboniferous age, 

 viz. in the Calciferous sandstone series of Scotland, it 

 occurs also in the ordinary Coal Measure nodules It 

 is in several respects more primitive than Lygmodeti- 

 dron, and in particular in the structure of its stele 

 comes nearer to that of ferns. The stele is, in fact, 

 a solid mass of primary wood and wood parenchyma, 

 corresponding in some degree to the protostele of a 

 simple type (see p. 61, fig. 36), but it has towards the 

 outside groups of protoxylem surrounded by wood in 

 both centripetal and centrifugal directions, which are 



Fig. 80. — Diagram of Transverse Sec- 

 tion of Lagenosfoma Seed near the Apex, 

 showing the nine flutings / of the coat 

 c\ J, the vascular strand in each; nc, 

 cone of nucellar tissue standing up in 

 the fluted apex of the nucellus n\ pc, 

 the pollen chamber with a few pollen 

 grains ; s, space between nucellus and 

 coat. Compare with diagram 56. 



