96 BOTANICAL GAZETTE [Avcust 
very complete historical account of the controversy on the morphology 
of the ovulate cone. He agrees with CELAKovskY that the arillus 
of Taxaceae, the ligule of Araucaria, and the ovuliferous scale of 
the other Pinaceae are homologous and are all a second integument, 
that all the ovulate cones are morphologically compound, but that the 
sporophyll is suppressed and represented by the ovule alone. 
THOMPSON (18) in 1909 brought out some interesting data in regard 
to the inversion and its relation to the theory of the axillary shoot. 
The inversion of the bundles supplying the ovule he explains as 
normal for sporangial supply, and cites cases of such inversion in the 
microsporophylls of cycads and of some conifers. In Tsuga he found 
two inversiens, the ovular supply being inverse to that of the scale, 
which in turn is inverse to that of the bract. The first inversion he 
considers as homologous with that of the scale of Saxegothaea and 
related to the ovule, while the second may be explained by the theory 
of the compound nature of the sporophyll. On this basis he makes 
two groups of conifers, the Araucarineae and Podocarpineae having 
simple strobili and ovules on the morphological upper surface of the 
scale, and the other Pinaceae having compound strobili and ovules 
morphologically abaxial. 
Tison regards the arillus of Saxegothaea and the ligule of Arav-_ 
caria as homologous with the ovuliferous scale of other conifers, 
calling them an ovuliferous appendage. He does not commit 
himself in regard to the axillary shoot theory as affecting Abietineae, 
Cupressineae, and Taxodineae, but agrees with THompson that the 
cones of Saxegothaea, Podocarpineae, and Araucarineae are simple. 
He favors the inclusion of Saxegothaea and Podocarpineae in the 
Araucariales suggested by Sewarp and Forp (14). 
CONCLUSION 
After reviewing the whole situation, one is impressed with the 
fact that both Podocarpineae and Araucarineae are very primitive; 
that they are probably related, but that the question is by no means 
settled. The whole coniter group still appears as a maze of cTos 
resemblances. If we confine ourselves to one or two charactets; 
the problems of relationship are comparatively simple, but bring 
more into consideration, and we are immediately in trouble. The 
Sacre aie Oi, ee wee 
