too Lady Isabel Browne. 
fertile there were two sporangia on each sporangiophore; while in 
S. Dawsoni each sporangiophore bore a single sporangium. In 
the ill-understood S. trichomatosum each leafy bract-like 
structure apparently bore a single sporangium. There is also 
considerable variety in the numerical proportion the ventral 
segments bear to the subtending dorsal ones. In Sphenophyllum 
fertile , where both dorsal and ventral segments are sporangiophores, 
they are equal in number. This is so also in Cheirostrobus, in which 
the dorsal lobes are sterile bracts, and if reduced “ sporangio¬ 
phores” are present in Sphenophyllum majus and 5. trichomatosum 
they are equal in number to the bracts, though unlike the latter 
they are undivided. In the types of S. Dawsoni and Bow- 
manites Romeri, however, the sporangiophores were twice as 
numerous as the dorsal lobes or bracts. If we accept, as the 
present writer feels compelled to do, Professor Lignier’s derivation 
of the vegetative leaves of the Sphenophyllales from a smaller 
number of more highly compound leaves by the separation round 
the stem of the lobes of the compound leaves, we shall be tempted 
to regard the quadri-sporangiate sporangiophores of Cheirostrobus , 
associated with an equality in number of the dorsal and ventral 
lobes of the sporophyll, as primitive. In that case the type of cone 
primitive for the Sphenophyllales had sporophylls divided into an 
equal number of dorsal and ventral segments, all of which were 
sporangiferous and bore four sporangia. No cone combining these 
characters is known ; but the occurrence of an equal number of 
dorsal and ventral segments in Cheirostrobus , the most ancient 
fructification of the Sphenophyllales known to us, and in Spheno¬ 
phyllum fertile, which has been shown to be probably very 
primitive, marks out this character as a primitive one. That the 
quadri-sporangiate condition is also primitive appears very probable 
from its occurrence in a type so widely separated from Cheiro¬ 
strobus as Sphenophyllum majus ; this is further supported by the 
prevalence of quadrisporangiate sporangiophores in the older 
Equisetales; this is of importance, since it is now generally 
admitted that the Equisetales and Sphenophyllales had a common 
origin from strobiloid forms (6), (7), (9), (4). On this theory the 
type of Sphenophyllum fertile was evolved from the above 
hypothetical ancestor by reduction of the sporangia on each 
sporangiophore to two ; in Cheirostrobus the sporangia were still 
formed in fours on the ventral segments, but the dorsal segments 
were converted into sterile bracts ; it seems likely that in Bow- 
