174 TREE ANCESTORS 



life history of a modem form lends to repeat the Hfe history of its 

 ancestral line. Starting wdth figure 9 the reader will note an 

 incipient lobe on either side of the base of an otherwise normally 

 shaped leaf. The larger of these two basal lobes, that on the right, 

 has a peculiar arrangement of the veins very similar to the arrange- 

 ment found in a modern stipule. The next stage may be illus- 

 trated by figure 4 in which these basal lobes have become 

 separate leaflets, like stipules but at the top instead of at the base 

 of the leaf-stalk. Or this second stage may be illustrated by 

 figure 3 where the leaf-stalk has not yet become elongated and 

 where on one side it is winged, much like the condition preserved 

 in the Upper Cretaceous form known as Liriodendron alatum (sho^\^l 

 in figure 4 on the plate of fossil leaves) and on the other side the 

 wing is separating from the rest of the blade in the direction of an 

 affair that looks like a modem stipule. By the elongation of the 

 leaf-stalk of a form like that shown in figure 3, a form like figure 1 

 would be approximated, figure 2 shows a succeeding stage, and it 

 needs but to separate the stipules from the leaf-stalk of a form 

 like that shown in figure 2 to get the normal condition of the mod- 

 em leaves such as is sho\^^l in figure 7. 



That the bud scales of the large blossoms have a similar origin 

 is shown by the frequent presence of a midrib or of a vestigial 

 base of a leaf-stalk between the scales and this is frequently elon- 

 gated as in figure 8 with an awn-like thickened tip. About as 

 frequently this elongation will expand into a tiny ovate leaf Hke 

 that shown in figure 6, and not rarely a larger and more normal 

 shaped leaf like that shown in figure 5 will be present. These may 

 be compared with the figures of the corresponding occurrences in 

 magnolia. Much the same sort of a history can be observed in the 

 buttonball (Platanus), and it would seem that these extraordinary 

 specimens of tulip-tree leaves illustrate the manner of origin of 

 the stipules during Upper Cretaceous times. 



The geological record of the tulip-tree is remarkable in several 

 ways, especially in the great variety and wide range of the Upper 

 Cretaceous forms. The oldest known are those of the Atlantic 

 border in New Jersey. This in itself is interesting because much 



