Berry: Origin of Stipules in Liriodendron 495 



existing species of Magnolia from our Southern States, Mag noli 



Fraseri Walt, and Magnolia macro ph) 



Magnolia 



is 



Amboy Clays of New Jersey 



which exhibits almost every stage from leaves with a truncate 

 base to those with an extremely lobed base. We find these forms 

 of especial interest because of the close relationship between Mag- 

 nolia and Liriodendron. The leaf figured on Plate 41, Fig. 9 rep- 

 resents the second stage of the series five-eighths natural size. 

 It is a very remarkable specimen, both in shape and venation. 

 Nearly triangular in outline, with an acute base, it bears upon the 

 petiole immediately below the base of the blade, but entirely 

 separated from it, a pair of appendages * shaped like flasks 

 with crooked necks (irregularly umbonate). The upper portion 

 of the appendages overlaps the basilar portion of the leaf, and like 

 it they are very coriaceous, thus differing widely from the thin 

 parenchyma of ordinary stipules. The length of the petiole be- 

 low the appendages is one and seven-eighths inches, showing that 

 they are true leaf segments and not of stipular origin. 



On Plate 42, Fig. r, is pictured a leaf which was borne on a 

 shoot from an old stump ; it is an inverted triangle in rough out- 

 line with a truncate apex four and three-quarter inches in width, 

 and tapers to a narrow base : the petiole is short and broadly 

 winged : these wings are continuous with the blade, there being 

 simply a contraction on one side, while on the other side there is 

 a sharp sinus which extends to within one-sixteenth of an inch of 

 the petiole, leaving room however for a plainly marked vein which 

 passes from one to the other : the wings are exactly similar to the 

 Wade proper in texture and extend to the base of the petiole, 

 encircling the stem as do true stipules ; the right hand wing is the 

 larger of the two and bears a mucronate point at its apex exactly 

 like a true stipule. While the latter might possibly be considered 

 by some as an adnate stipule, the left-hand wing can hardly be 

 considered other than as a winged margin of the petiole in the act 

 of becoming stipular, and rather closely resembles Liriodendron 

 alatnm Newb. from the Laramie. 



The left hand one is broken off in the specimen, but the scar where it was at 



■ 



tached is plainly shown. 



/ 



