Humphreys: An analogy 575 



is different from his, which is that the lobed forms are the primi- 

 tive or ancestral type, and is, I think, more in accord with the 

 recapitulation theory. Indeed, Jackson recognizes the difficulties 

 of his interpretation and that under it Sassafras becomes an ex- 

 ception to the usual condition. He says :* "The fact that seed- 

 lings start with entire leaves and later acquire lobed leaves like the 

 earliest fossil representatives seems difficult to harmonize with the 

 usual condition, where the first leaves are like the primitive or 

 ancient types, and later leaves are different, being more special- 

 ized, as in Liriodendron, white ash, and Plat anus" Now while 

 it is true that what are generally recognized as the earliest an- 

 cestral forms of Sassafras leaves are lobed, and that apparently 

 there are no entire fossil Sassafras leaves known, as remarked by 

 E. W. Berry in his " Notes on Sassafras/' f nevertheless, as the 

 latter author well says, there are several fossil leaves with entire 

 margins, which are referred to other genera, that might with 

 equal propriety be referred to Sassafras, such for example as 

 Cinnamomum Heerii Lesq.,* and anyone familiar with figures of 

 fossil leaves may readily recall other species in this and other 

 genera which compare very closely with certain of the entire leaf 

 forms in the living Sassafras. It may also be pertinent to remark 

 that the determination of genera in fossil botany, based upon leaves 

 only, has not attained to such precision that all leaves can with 

 certainty be referred to their correct genera. Further than this, 

 it may be recalled that even the correctness of the reference of 

 many fossil lobed leaves to the genus Sassafras has been ques- 

 tioned or criticized by competent authorities. Thus the fact that 

 lobed leaves were actually the earliest ancestral forms is by no 

 means definitely established. The entire forms may yet be found, 

 either as new discoveries or as species incorrectly referred to other 

 genera. Jackson's interpretation, therefore, which places the 

 lobed leaves as the earliest forms, may not be in accordance with 

 the actual facts, but opposed to them, and since it necessitates an 

 apparent exception to the usual condition, requiring a devious 

 explanation, the better course would seem to be to regard the 



* Loc. cit. 1 08. 



tBot. Gaz. 34: 426-450. //. /8+f. r-4- x 9°2 



X Loc. cit. 433. 



