430 



BOTANICAL GAZETTE 



[DECEMBER 



and in several instances are even retuse. The base of the leaf is 

 in all cases decurrent and the angle of divergence of the ascend- 

 ing margins never exceeds and seldom equals 45°. While the 

 comparative size of the lobes varies somewhat, the leaves are 

 approximately bilaterally symmetrical. This form seems to have 



escaped attention hereto- 

 fore, and none of the fos- 

 sil forms referred to this 

 genus resemble it in the 

 least except S. platanoides. 

 An ancient form which re- 

 sembles these five-lobed 

 leaves in outline is Fon- 

 taine's Ar all ccphy Hunt obtit- 

 silobiim ; the latter, how- 

 ever, has thicker veins and 

 the lateral lobes are not so 

 extensively developed ; and 

 the lateral which runs from 



the 



P 



rim 



ary 



to the extra 



lobe is inserted much nearer 

 the base than in the mod- 



FiG. 4. — Abnormal leaf of Sassafras; ern leaves, 

 reduced two-thirds. t-, . ■ r i.\^^ .^r^A 



The venation of the mod- 

 ern leaf is uniform in its variation. Safe rules for guidance 

 in determining fossil forms would be the following: The prima- 

 ries are opposite or sub-opposite (while this is not strictly 

 applicable to the simple leaves it seems to be general in the 

 lobed ones); they branch from the midrib a considerable dis- 

 tance above its base, thus differing from Aralia, Cissites, 



Platanus, etc.; the base is never enlarged to form basal lobes 

 as it is in Menispermites, Araliaephyllum, Protophyllum, Aspidio- 



phyllum, Platanus basilobata, etc ; margins are entire ; obviously, 

 if any of the ancestral forms were dentate, this character would 

 reappear occasionally in their descendants (all known Laura- 

 ceae have entire margins). A somewhat constant character 

 in Sassafras leaves is the venation of the basal portion of 



