34 



species of Liriodendron and^I will not attempt anything further 

 in that line at this time, but will confine myself in these notes to 

 calling attention to several "interesting forms of these leaves and 

 briefly discussing the evidence they offer as to the probable rela- 

 tions of some of the ancient members of this genus. 



While Liriodendron Tulipifera has long been known to have 

 variable leaves the extent of this variability has not been dreamed 

 of, nor any reason assigned which would account for it. Dar- 

 win's law that wide-ranging species are variable is fully carried 

 out, but, on the other hand, Sedgwick's rule that old species 

 have lost their variability is not fulfilled. In fact, quite the reverse 

 holds good, Liriodendron having reached quite a respectable old 

 age and still retaining its variability with all the vigor of its 

 Cretaceous days. 



The accompanying plates~picture some especially interesting 

 leaves, all one-fourth natural size, from a collection of several 

 hundred. One of the most curious is the sport shown at Fig. 7. 

 The left half of the blade is somewhat normal in shape but the 

 right half is reduced to a narrow lanceolate strip, which in vena- 

 tion bears a striking resemblance to an ordinary cotyledon. We 

 have reconstructed this leaf, Fig. 6, as if both halves of the 

 blade were narrowed as is the right half; this gives us a leaf 

 strikingly like what we consider the primitive Liriodendron of 

 the early Cretaceous or Jura-Cretaceous to have been. For 

 comparison we show an ordinary cotyledon at Fig. 8, which, as 

 will be seen, is very similar to Fig. 6. 



The original Driodendron leaf was long and narrow and as 

 time passed there was a progressive widening of the blade and a 

 corresponding reduction of the apex. It has been suggested 

 that the mucronate point which usually tips the midrib of the 

 modern leaf is a surviving rudiment of this once pointed apex. 

 However this may be, we often find leaves with the acute ances- 

 tral apex (Figs. 4, 11, 12, 13, 14, 15). While the leaves bearing 

 the tips shown at Fig. 1 1 were otherwise normally shaped leaves 

 of large size, and while the leaves shown at Figs. 14, 15 were 

 otherwise normal, the remaining acute-tipped leaves are very sug- 

 gestive. The leaf shown at Fig. 4 is almost identical with the 



