■0 

 1858.] PHILLIPS FOSSIL FRUIT. 49 



phuret of iron. In it, near the base, deep grooves appear (corre- 

 sponding to the external ribs) ; two of these are very ^stinct, and, 

 after passing less than half the length of the interior, cease symme- 

 trically. 



By the same fracture the thickness of the shell or crust is disco- 

 vered, varying from ^th to y^^^ ^^ ^^ inch. The whole of this 

 substance is fibrous. 



In a geological point of view, the more precise determination of 

 several of our fossil fruits appears desirable ; especially of the Oolitic 

 and Wealden periods. But this is far from easy. If, for example, 

 we endeavour to trace the analogies of the present specimen, and 

 begin with the great Monocotyledonous groups, we may regard its 

 fibrous mass as having a resemblance to the fibrous pericarp of the 

 Palmacece. But what Palm gives us a regularly eight-ribbed fruit ? 

 What Monocotyledon exhibits such remarkable deviation from the 

 usual tendency to be represented in threes ? 



Without stopping among the Coniferce and Cycadaceoe, which offer 

 in their fruits nothing of this kind for comparison, we may enter 

 the Dicotyledonous groups. Keeping in view the probability of the 

 object being of a firm or woody texture, like a nut, we find in the 

 Juglandacece some analogies worth notice. For example, the Hickory- 

 nuts have often four stronger meridional ridges alternating with 

 four weaker elevations, corresponding to the quadrilocular interior 

 cavity. The outer covering is four-ridged in Carya amara. Ranged 

 among the JEwphorbiacece, in the new Museum at Kew, I find an 

 example of a quadrilocular capsule, — trilocular capsules being com- 

 mon in this order. The plant is from North Australia, and is named 

 Petalostigma quadrilocularis. As far as general appearance and 

 character of figure may be appealed to, this fruit is the most like 

 my Swanage fossil. The specimens at Kew show the pericarp to be 

 marked by eight meridional ridges, while the interior divisions of 

 the fruit correspond in number, and separate in octants. 



It may, perhaps, here deserve notice, that a spheroidal many- 

 ribbed body, which is usually regarded as a fruit, and was obtained 

 from the lowest part of the Oolitic series on the Yorkshire coast, at 

 Whitby, has been compared with Hura crepitans, the Sand-box-nut 

 of the West Indies, which is also Euphorbiaceous (see Young and 

 Bird, Geological Survey of Yorkshire Coast, pi. 1. fig. 5, ed. i., pi. 2. 

 fig. 6, ed. ii.). 



I purposely abstain from recording the ineffectual attempts I have 

 made, by my own inquiries, and by consultation with others, to settle 

 the place of this fruit among the natural orders of plants. Probably 

 further search in the same locality will yield additional data, and 

 render it easy to determine what is now found hard to guess. 



VOL. XV. PART I. 



