i 
Wate es Sheed Tote 
Oe 
Rae Poe 
1909] COULTER—EVOLUTION AMONG GYMNOSPERMS 87 
by the shortening of the sporophyll-bearing shoot is a conspicuous 
feature of gymnosperms, and it must have been derived from the 
condition observed among the Cycadofilicales. 
The Cordaitales were the first gymnosperms to produce strobili, 
and this is one of their conspicuous contrasts with Cycadofilicales. 
The record of the structure of their strobili is meager, but it shows 
several tendencies in strobilus-formation. Of primary importance 
is the fact that the strobili are monosporangiate, and this monospo- 
rangiate character prevails throughout the Ginkgoales and Coniferales. 
Among the Gnetales, a group probably related to the general conifer- 
ophyte phylum, amphisporangiate strobili occur in Tumboa. If 
this connection be accepted, therefore, these amphisporangiate 
strobili have been derived from monosporangiate strobili. It is not 
necessary to associate in one genetic connection all of the amphispo- 
rangiate gymnosperms, for that condition doubtless appeared inde- 
pendently several times, just as the monosporangiate strobilus is 
known to have appeared at least twice in distinct phyla (Cordaitales 
and Cycadophytes). 
Another fact in reference to the strobili of Cordaitales, which 
must stand for the most ancient gymnosperm strobili, is that they 
included both simple and compound strobili. The staminate strobilus 
was simple, that is, its sporophylls were borne directly upon the axis 
of the strobilus; and this type of staminate strobilus persisted through- 
out the Ginkgoales and Coniferales. Among the Gnetales the 
staminate strobilus is compound, the individual simple strobilus 
being borne on axes of the second order in the axils of sterile bracts 
which make up the general strobilus. There is an evident relation-~ 
ship between the compact compound staminate strobilus, such as 
occurs in Ephedra and in Tumboa, and the short foliage branch 
bearing axillary simple staminate strobili, as in Torreya. ‘Even in 
Gnetum the compound staminate strobilus is a loose one; and among 
the taxads there is a tendency to compact the staminate strobiliferous 
branch. The conclusion is that the staminate strobilus among coni- 
ferophytes is quite persistently simple, but that in the more modern 
members of the phylum it tends to become compound, a condition 
accomplished by compacting a short strobiliferous shoot. 
The ovulate strobilus of Cordaitales was compound, at least in 
