Johnson — Ginkgophyllum kiltorkense. 173 



go. The region in the lamina where the lobes meet is ordinary mesopliyll 

 tissue, free from veins. Further, the veins of one lobe do not unite with those 

 of the adjoining one for some distance below the point of separation of the 

 lobes. Thus the segmentation might be carried deeper and deeper until the 

 point of junction of the veins, well towards the base of the leaf, is reached, 

 giving a leaf-blade comparable to that of Baiera, with its numerous digitate 

 segments, or conversely, tiie mesophyli might be imagined as filling up the 

 free space between the segments, gradually obliterating them, until the lamina 

 would assume the condition of the more or less bilobed leaf of Ginkgo itself. 

 In pursuit of this idea I took a leaf of Ginkgo, and cut through the mesophyli 

 between the forking-veins from the broad apex downwards, until the points of 

 fusion or bifurcation of the veins were reached. The result, showing six main 

 artificial lobes, is illustrated in PI. XI, fig. 1. Its suggestion of the fossil, 

 Ginkgo digitata, or of Baiera (10), or of the Kiltorcan specimevi, is sufficiently 

 obvious. The text-books give illustrations of abnormal leaves of Ginkgo 

 deeply, occasionally palmately lobed. More interesting is the fact that one 

 cotyledon of the Ginkgo seedling (wlien not both) is deeply bilobed, as are 

 normally the first leaves of the seedling. A plant's embryonic leaves reveal 

 the ancestral characters of its group ; and those in Ginkgo show a closer 

 connection with Ginkgophyllum than is evident from the examination of its 

 adult leaves. " Pro-Ginkgo," in becoming " Eu-Ginkgo," iu the course of 

 time, has filled in the mesophyli, and has thus obliterated the lobes more or 

 less completely. Its vascular bundles have become more numerous, too, by 

 more frequent forking. 



The disappearance of lobation is in keeping with the general conclusion of 

 fossil evidence — viz., that the earlier types (Baiera) show more deeply lobed 

 leaves, and that, as later Mesozoic passes into Tertiary, the normal wedge- 

 shaped, simply lobed, or unlobed, long-stalked leaf, characteristic of the 

 extant Ginkgo, to the final exclusion of the segmented type, predominates. 



Still, with the warning furnished by the errors of the past iu identifying 

 as ferns many of the fossil j'lants of the Carboniferous epoch, now known to 

 be Pteridosperms, by relying maiuly on conclusions drawn from their foliage 

 and its venation, it may seem rash to conclude from the leaf alone that the 

 Kiltorcan specimen is undoubtedly a Ginkgophyte. 



My own impression is that it is a true Ginkgo ancestor ; and iu consequence 

 I prefer to use for it the generic name Ginkgophyllum, Saporta, rather than 

 the name Psygmophyllum, Schimper, which, while less suggestive of pre- 

 judging than Saporta's name, is also less indicative and is better applied to 

 the less segmented form of leaves, such as Psygmophyllum WiUiamsoni may 



