Sphenopteris Type of Foliage. 59 
Usually each of the segments bears a single seed, though more rarely they 
appear to have bifurcated, each lobe terminating in a seed. 
The Seeds. 
The seeds themselves are extraordinarily small objects. Two 
examples are figured on Plate VI, Figs. 2 and 3, magnified twenty times. 
Their average length is about a millimetre, the largest being 1-2 mm., 
while at their greatest width they measure *75 mm. across. They seem 
to be of the radially symmetrical type. They are more or less oval in 
shape, and exhibit several rather sharp, longitudinal ridges (PL VI, Fig. 3). 
They do not appear to show any special apical peculiarities. 
There would seem to be some grounds for the belief that most of the 
seeds seen on this specimen are enveloped in closely-fitting cupules 
(Plate VI, Figs. 4-8). It is, however, very difficult to decide whether such 
an organ is really present, for not only are the seeds very minute, but the 
cupules, if they exist, are closely similar to them in size and shape. In 
such impressions as these, it must always be difficult to demonstrate the 
presence of a cupule, unless it has, as in the case of Lagenostoma Sinclairi , 
Kidst. MS. 1 , a form quite different from that of the seed itself, or unless it is 
a deeply cleft structure, as for instance in Calymmatotheca Stangeri , Stur 2 . 
In the present case, if cupules are present, they appear to correspond more 
closely to those of Lagenostoma Lomaxi *, though they are perhaps less 
deeply divided, and without the glandular structures found on the cupule 
of that seed. Impressions of such cupules must naturally be difficult 
to distinguish from those of the seeds themselves. 
In this specimen, there do not appear to be any good cases in which 
empty cupules, without seeds, can be recognized. On PI. VI, Fig. 2, 
there is a somewhat indefinite body, seen to the right of the naked seed, 
which may be of this nature, but it is impossible to determine its precise 
nature. 
In many instances, however (PI. VI, Figs. 5-8), there are indications 
which suggest the presence of an outer investment containing a seed, the 
surface of which does not appear to correspond exactly with that of the 
naked seed. In such cases, the shape of the supposed cupule differs 
somewhat from that of the seed itself, and its length and breadth are slightly 
greater. The longitudinal ridges are also less prominent. Here and there, 
examples (PI. VI, Figs. 5 and 6) may be found, which appear to show that 
the outer investment or cupule may have been lobed, though at this stage 
of development the lobes remained appressed to the seed, and were not 
1 Arber (’05 1 ). 
2 Stur (77), p. 151, PI. VIII (XXV), Figs. 5-7, and Text-fig. 27 on p. 158. 
8 Oliver and Scott (’04), p. 217, Text-fig. 2. 
