502 Agnes Arber.—On the Structure of the Palaeozoic Seed 
The ridge, unlike that in Cardiocarpum acatinn} only extended a certain 
distance up the shell and then died out. There is a section 1 2 * from Oldham 
in the Williamson Collection, which is referred in the MS. catalogue to 
Cardiocarpon anomalum , but which is in reality a most characteristic 
example of his C. compression . It is a transverse section showing a well- 
marked median ridge (Text-Fig. 2,c). It was probably this ridge which 
deterred Williamson from referring the section to C. compression. The 
attribution to C. anomalum is a curious one, since all the other sections in 
the Collection, referred by Williamson to Carruthers’ species C. anomalum? 
have proved to be Lepidocarpons. 4 
A large number of seed impressions of the Cardiocarpon type from 
Devonian and Carboniferous rocks have been figured by various authors, 5 
but none of them can with certainty be identified with Mitrospermum 
compression. 
B. Generic Name. 
Brongniart’s original description of the genus Cardiocarpon 6 dealt 
only with external characters, and was somewhat vague. It ran as 
follows :— 
‘ Cardiocarpon. Fruits comprimes, lenticulaires, cordiformes ou reni- 
formes, termines par une pointe peu aigue/ 7 
Since Brongniart’s time various seeds, which he would probably have 
included in the genus Cardiocarpon , have been distributed into several newer 
genera, e. g. Samaropsis, Cordaicarpus , Cyclccarpus , &c. A critical dis¬ 
cussion of these genera, which are distinguished by slight differences in 
external form, has been published by Kidston, 8 who concludes that it is 
advisable tentatively to retain Brongniart’s generic name Cardiocarpon for 
the whole assemblage of these seeds. 
Bertrand 9 has recently published a fresh description of the genus, 
founded on a re-examination of the silicified seeds previously studied by 
Brongniart and Renault, in which he discusses internal as well as external 
characters. 
On comparing the structure of Mitrospermum compression, as outlined 
in the preceding sections of this paper, with the generic characters enumerated 
by Bertrand for Cardiocarpus , we find that they fail to agree in one important 
point, namely, the type of vascular system. Bertrand restricts the use of 
the name Cardiocarpus to forms in which the hilo-chalazal bundle emits its 
1 Lindley and Hutton (’ 31 ), PI. LXXVI. 2 W. 1417. 
3 Carruthers (’ 72 ). 4 Scott (’01). 
5 Dawson (’ 71 ), Newberry (’ 73 ), Andrews (’ 75 ), Kidston (’ 94 ), &c. 
G The name has been used by different writers in the forms :— Cardiocarpon [Brongniart (’ 28 ), 
Williamson (’ 77 )], Cardiocarpus [Brongniart (’ 81 ), Bertrand (’ 08 )], and Cardiocarpum [Lindley 
and Hutton (’ 31 )]. Compare Scott and Maslen (’ 07 ), footnote to p. 90. 
7 Brongniart (’ 28 ). 8 Kidston (’ 94 ), p. 263. 9 Bertrand (’ 08 2 ). 
