British Coal Measures. 
83 
I do not propose in the present instance to deal at all with structure 
specimens, still less with the difficult correlation of such petrifactions with 
impressions. My present task is merely to sort out and to reclassify the 
impressions. 
From the point of view of the study of impressions the testa and 
its characters are all important. I assume that very few radiospermic 
seeds, as Oliver 1 has pointed out, and as experience with impressions con- 
firms, have a testa differentiated into a fleshy outer sarcotesta and a hard 
inner sclerotesta. The converse, however, does not hold good. Some 
Platysperms, such as Cardiocarpus , as here restricted, Cor?iuspermum 
and Cordaicarpus , show no signs of a differentiated testa. In others, how- 
ever, such as Samaropsis and Samarospermum , the differentiation of the 
testa is very obvious. In the latter type, it is convenient to speak of 
the central portion of the seed as the ‘ nucule ’, an ancient term implying 
the nucellus surrounded by the hard sclerotesta. The soft sarcotesta as viewed 
in impressions is nearly always flattened over the nucule, and the rim 
of sarcotesta surrounding the nucule is conveniently described as the 
‘ wing ’. 
I regret that it has been found necessary here to institute a large 
number of new generic terms, partly because a considerable number of 
new species are described for the first time, but more so because there is 
a very great variety among seed impressions, — a variety which must find 
adequate expression unless these fossils are to remain bunched indis- 
criminately into such nondescript genera as Carpolithus . On the other 
hand, I am well aware that it may eventually be found that more than 
one generic type has here been included in a single genus, especially in 
the case of the new genus Radiospermum. But this is a defect which only 
time and further evidence can remedy. The total number of genera here 
recognized is 14, and 37 species are described. The latter part of the paper 
is devoted to an attempt to define these genera and species more exactly 
than has hitherto been possible. Of the 14 genera here recognized, 9 are 
new, the remainder have been already employed for continental specimens. 
On the other hand, of Dr. Kidston’s 18 species, all but 4 are recognized, 
though the great majority of them are referred to other genera. Ten new 
species are described here, and twenty-seven others are recognized which 
have either been previously recorded by Dr. Kidston, or have been unknown 
from Britain though known from the Continent. A figure of every British 
species is included. 
It is hardly necessary to point out perhaps that the detached seed 
impressions, here under consideration, belong to at least two very diverse 
groups, the Pteridospermeae and the Cordaitales. At present we are 
unable to discriminate between the seeds of these groups. All that we can 
1 Oliver (’OB), p. 453. 
