GYMNOSPERM^E. 71 



Arctica,' vol. ii, p. 478, pi. li, figs. 2 and 3, reprinted from the ' Phil. Trans.,' 1869), bear a 

 striking resemblance to the Ballypalady Pine cones, while possessing little or no structure 

 to support a contrary view. 



The fruit of Magnolia consists of a number of carpels or follicles in a compact spike, 

 which open along their outer edge to allow of the escape of the seed. In Magnolia 

 grandiflora, to which the fossil is compared, the deciduous sepals and petals leave deep 

 scars in falling away, forming a characteristic region on the stem immediately beneath 

 the fruit, and the seeds are arranged not spirally, but horizontally in whorls, at right 

 angles to the axis. The characteristic impress of a Magnolia fruit, whether the carpels 

 were dehiscent or not at the time, would be the external, and not an internal cast, 

 such as those met with in Greenland. Heer, in describing them, only says that the cones 

 are "much resembling those of Magnolia," and that several of the "carpels show the 

 fissure, precisely as in Magnolia glauca," The supposed fruits were not found in the 

 same locality with the supposed leaves. It is easy to see how the division between two 

 pine seeds might be mistaken for a fissure, and it is not impossible that the previous deter- 

 mination of Magnolia leaves may have led to an identification that might otherwise not 

 have suggested itself. The specimens were unfortunately not returned, and the question 

 of their true nature therefore remains open, — a fact to be regretted, since much stress had 

 been laid on their presence to support the rather excessive estimate of the mildness of 

 climate in high latitudes during the Tertiary period. 



Pinus Plutonis bears the greatest resemblance to the less hardy of the living species 

 of Pine. Pinus sinensis, of Hong Kong, differs chiefly in its slightly smaller cones and 

 seeds, and more slender needles. Pinus laricio, an inhabitant of the Mediterranean and of 

 Western Asia, varies in the more pyramidal form of its cones and the greater size of the 

 heads of its scales, the needles and the scars left by them on the branchlets being nearly 

 identical. Pinus halepensis is identical with it in the size and ornamentation of its cones 

 and scales ; but, while our cones are less pyramidal and apparently grow in an upward 

 direction (PL XV, fig. 3), those of P. halepensis grow in pairs inversely downwards. The 

 seed-wings of the fossil (PI. XVII, figs. 2, 3, 8), are more rectangular and less obliquely 

 truncated at the extremity than are any of those of the existing Pines which otherwise 

 resemble it. The needles are very much the same as those of P. halepensis, and 

 perhaps a trifle more slender than those of P. laricio. Our fossil must thus be regarded 

 as a different, but closely allied species, and requiring probably a similar mean temperature. 



P. halepensis is usually a low-spreading tree in habit, from twenty to thirty feet in 

 height, though some of its varieties form much loftier trees, and ripens its cones in 

 the autumn of the second year. It is common in the Mediterranean region of Europe 

 and Asia Minor, but is not found north of the Apennines nor at elevations above 2000 feet. 



Ettingshausen has endeavoured to trace a genetic connection between the Pines of 

 the various Tertiary formations of Austria, especially of Leoben in Styria 1 and existing 

 1 ' Beitr. zur Erforschung der Phylog. der Pilanzelemeuten,' 18/7. 



