oblong sunk, continued by a minute short oblique fissure at the outer corner near the 

 raphe, and by a similar fissure at the inner corner at the opposite end of the hilum. 



Length 11 mm., breadth 12 mm., thickness 6 mm. Reuver. 



This large hard plano-convex seed certainly belongs to the Sapotaceae, and shows 

 a great resemblance to Mimusops browniana Benth. a tree of Queensland, but our fossil 

 is considerably smaller. It does not show so close a resemblance to any other species 

 with which we have been able to compare it, for most of the seeds of Sapotaceae differ 

 in shape and are attached obliquely. We figure Mimusops browniana (figs. 30a, b) not 

 only for its general resemblance, but as showing the same small oblique folds which 

 we notice in our fossil. What is the origin of the folds we cannot say without recent 

 material for cutting sections. Our fossil is unique and we cannot dissect it. The folds 

 look like ..germination grooves", to assist the splitting of the very hard thick seed. 



The genus Mimusops is now essentially tropical, as are nearly all the Sapotaceae. 



EBENACEAE. 



DIOSPYROS LOTUS Linn. 

 PI. XVI, figs. 1,2. 



Kernel large semiovate laterally compressed; surface covered with small hexag* 

 onal pits. 



Length 13.0 mm., breadth 5.5 mm. Brunssum. 



This seed agrees in size, shape, attachment, and surface sculpture with that of 

 D. Lotus, a tree wild in temperate Eastern Asia up to 1300 metres in China, and culti* 

 vated in the Mediterranean Region. We have only one fossil seed. 



It is worth noting that when we first compared this seed with that of D. Lotus, 

 though the agreement in size, shape, and method of attachment was so striking, we 

 rejected the identification, because the surface sculptures were so different. The surface 

 of the testa is very peculiar in D. Lotus, resembling that oijasmimum. This layer of long 

 shining zig-zag cells (fig. 2) is easily rubbed off when macerated, leaving a second layer 

 of zig*zag cells. This too can be removed, exposing a hard bony kernel with hexagonal 

 pits, exactly as in our fossil (fig. 1). We mention this as a striking instance of the 

 necessity, when comparing recent and fossil fruits and seeds, of making certain that we 

 are comparing similar coats. Our fossils have been macerated; before comparison it is 

 therefore necessary to macerate the recent specimens also. 



DIOSPYROS Sp. 2. 

 PL XVI, fig. 3. 



Kernel ovate roundly pointed above and below, ventral edge gently curved, 

 dorsal edge strongly curved; attachment prominent bent towards the ventral side; 

 surface much wrinkled longitudinally and covered with minute pits. 



Length 9 mm., breadth 4 mm. Swalmen. 



130 



