GYMNOSPEKMS-OORDAITALES— (JORDAITES, 259 



the salient points of more general interest may be mentioned the absence 

 of primary wood in the trunks, the presence of both the primary and 

 secondary development in the roots, while the vascular strands of the 

 nerves in the leaves comprise a triangular primary axis partially surrounded 

 by a secondary growth. The trees of the Corclaitece grew rapidly to a con- 

 siderable height, branching only near the top. The flowers were diclinous 

 and aperianthous, but whether the unisexual flowers were monoecious or 

 dioecious is not yet known. Both sorts were mingled in the fossil state. 

 The female flowers are monocarpal, for although, like the male flowers, they 

 are cone shaped in general aspect, they are solitary, each female flower 

 being surrounded by an involucre of bracts. The male flowers are in small 

 cones spirally arranged in the axils of bracts about a rather robust axis. 

 Each flower is composed of two or three stamens, each compiised of a fila- 

 ment bearing three or four longitudinally dehiscent anthers, which are free 

 above and united at their bases. The pollen grains are ellipsoidal in section 

 and very abundant. In the anther of one species the grain measures dOju in 

 longer and bOju in shorter diameter, while in the pollen chamber, which is 

 constantly present in and forms an interesting feature of the seed, it measures 

 121/^ and 72 ju, respectively. The seeds, including among others the Cordai- 

 carjms, have two envelopes. The external envelope (Sarcotest) is fleshy and 

 is sometimes traversed by elongated fibrous cells mingled with canals contain- 

 ing gum or tannin. The internal covering (Endotest) is formed of densely 

 lignified cells and suggests the shell of a nut. The ovules are orthotropous 

 and erect. A pollen chamber, relatively little developed, is always found 

 in the summit of the nucleus, and the pollenic canal is always attached 

 to the micropylar tube of the outer envelopes. The archegonia are in 

 pairs. No embryo has yet been found in any of the fruits, although the 

 latter appear to have been fully developed. Renault points out that in 

 Cordaites, as in the living Cemtozamia, the embryo was probably not devel- 

 oped until the seed had been placed some time in the soil. To the Cordaites 

 Renault seems to refer the fossil seeds which are rather flat and bilaterally 

 symmetrical. In his Com-s de Botanique Fossile^ he refers to the Cordaitece 

 the genera Cardiocarpus, Biplotesta, Sarcotaxus, Leptocaryon, Taxospermum, 

 and Bliahdocarpus, while M. Grand 'Enry has since" included Hypsihcarpus, 

 Cydocarpm, and Samaropsis in the same category. 



1 Vol. i, p. 102. ^ G^ol. et pal. bassiu houill. Gard, 1890, p. 312. 



