1082 
Thompson. — ■ The Anatomy and 
instead of two in each group. This is no doubt a specialized feature, for in 
the seedling (see PL XCVI, Fig. 27) one finds the simpler type of organization. 
The observations of former investigators differ considerably from those 
just outlined, and also differ materially from each other. Geyler, 1 the 
pioneer in this sort of work, observed the girdle at the node in E. equiseti- 
formis , but did not consider it of importance. He believed that the 
bundles were continuous through two internodes, the leaf-traces attaching 
themselves laterally at that distance below their exit. In reality they run 
through only one internode where they are formed from the girdle. He 
also made the mistake of considering the large internodal bundles as the 
leaf-trace bundles of the succeeding node. Van Tieghem 2 observed the 
intercalary bundle between the leaf-trace strands which Geyler failed to see. 
He thought that it was derived from the traces and divided into two to give 
the proper number of bundles for the next internode. Strasburger, 3 while 
considering that the bundles ran through two internodes, correctly observed 
the origin of the intercalary bundle from the girdle. All these investigators 
examined species, E. distachya , vulgaris , and equisetiformis , which belong 
to the simplest type described. As we have seen, the conditions existing in 
many other species are quite different, although intermediate stages exist. 
In spite of the disagreement in regard to details the course of specializa- 
tion may readily be traced. There are indications that in the ancestral 
form the internodal bundles were as in the simpler type, and the small 
bundles ran out intact to form the leaf-traces. The intercalary bundle 
was absent. From this condition the girdle increased, the intercalary 
bundle developed, and the traces lost their connexion with the internodal 
bundles. Finally, the third small bundle of the internode developed, and 
the traces came to alternate with the internodal bundles. 
Some of the features just described are of importance from the stand- 
point of relationships. The small number and simple regular course of the 
bundles through the internode are quite different from the conditions in the 
Cycads and Bennettitales, with which there has recently been a tendency to 
connect the Gnetales. It is well known that all the Cycadophyta had their 
primary vascular system arranged in the form of an irregular trellis from 
which the leaf-traces departed in a haphazard manner. The conditions in 
Ephedra resemble much more closely those of the Coniferales and Dico- 
tyledones. 
The double leaf-trace is a feature on which much stress has been laid. 
There appears to be general agreement as to its primitive character, for it 
is present in the ancient Cycadofilicinean forms. In Ephedra it differs from 
the double trace of many Gymnosperms in that the two strands of which it 
1 Geiassbiindelverlauf und Laubblattregion d. Coniferen. Pringsheim’s Jahrbiicher, 1867. 
2 Anatomie des Fleurs des Gymnospermes. Ann. de Sci. Nat., Bot., 1889.- 
3 Die Coniferen und die Gnetaceen, 1872. 
