2 34 
Review. 
In the detailed account of Cycadeoidea ( Bennettites ) Gibsoniana, 
phloem is doubtless a lapsus calami for periderm (p. 388, line 16).; 
internal phloem would be a remarkable feature in a Bennettitean 
stem. 
The author is sceptical as to the flower of C. Gibsoniana having 
been bisexual, for he finds no trace of the stamens (p. 395). 
Considering the proved bisexual character of closely similar 
American specimens, it seems most probable that the disappearance 
of the stamens is merely due to the maturity of the fruit, but here 
again Wieland’s later work must be consulted. 
On the vexed question of the morphology of the Bennettitean 
gynseceum the author takes the view that the seed-stalks (mega- 
sporophylls) and scales are homologous, being respectively fertile 
and sterile carpels, enclosed by a perianth of bracts (p. 403). 
Vectia, a large mass of fossilised phloem, discovered by Dr. 
Stopes at Luccomb Chine, is referred on good grounds to the 
Bennettitales, rather than to the Coniferae. 
Carruthers instituted a special tribe WiHiamsoniese, and it is 
probable that we shall have to return to this course, though in the 
meantime it may be difficult to draw the line between the William- 
sonian and Bennettitean tribes. 
Much attention is given to the difficult interpretation of the 
original specimens of Williamsonia gigas. There is room for some 
doubt as to the details, but the fact that Williamsonia flowers and 
Zamites leaves were borne on the same stem, confirming William¬ 
son’s well-known restoration, is fully established (see Pig. 541). 
The author maintains that the flowers of W. gigas and other 
species were bisexual, and that the funnel-shaped appendage 
believed to have surmounted the top of the gynseceum was identi¬ 
cal with a whorl of connate stamens. The complete flower, on 
this view, must have been, as Mr. Thomas puts it, top-heavy, for 
the staminate whorls were 3 or 4 inches across, or more when 
expanded. A more serious objection is that in all flowers of 
Bennettitales which are known to have been bisexual, the staminate 
whorl was hypogynous, below and exterior to the gynseceum. This 
is the case in Cycadeoidea , Cycadella, Wielaudiella and William- 
soniella. On Prof. Seward’s view, precisely the opposite arrangement 
prevailed in Williamsonia. The analogy with the hypogynous and 
epigynous flowers of Ericacese (p. 458) is scarcely adequate, for 
even in an epigynous flower the gynseceum is central, whereas on 
this hypothesis the central organ of the flower in the Bennettitales 
would have been sometimes the gynseceum, sometimes the and- 
rcecium. The comparison with Willi am soniella, so admirably 
investigated by Mr. Thomas, seems specially unfavourable to the 
author’s view, for here the agreement with Williamsonia is close, 
extending even to the presence of a “ corona ” at the apex of the 
female receptacle, and yet the staminate whorl is hypogynous. 
In this remarkable family all things are possible, but a good deal 
more evidence will be required before the author’s daring hypothesis 
can be accepted. 
Apart from controversial points, the whole account of the 
Williamsonia group is most interesting, and gives a striking idea 
of the wide range of organisation in the Bennettitales. Some of 
