TORREYA 



May, igio 

 Vol. lo No. 5 



VARIATION AMONG NON-LOBED SASSAFRAS library 



LEAVES * '^^^ ^^^^ 



By Edwin W. Humphreys OAl^OEf^ 



Sassafras is for various reasons one of the most interesting of 

 our native trees. One of the interesting features, and the one 

 that has probably attracted the most attention to the tree, is its 

 variously shaped leaves. That each of these leaf forms in turn 

 shows considerable variation in its characters is apparently not so 

 definitely known, and the limits of such variation are still less 

 known. Yet a study of these differences is of much interest, 

 particularly from the viewpoint of paleobotany ; for most of the 

 identifications of fossil plants are based on leaves only and natu- 

 rally the limits of leaf variation are of more importance to the 

 paleobotanist than they are to the botanist who has, in addition 

 to the leaves, other characters on which to base his identifica- 

 tions. It was for the purpose of determining, in a measure, how 

 greatly the non-lobed sassafras leaves varied among themselves 

 that the present study, based on leaves collected at random in 

 Bronx and Pelhani Bay Parks, New York, and on the Palisades 

 of New Jersey, was made. 



The most obvious variation is in the proportion of length to 

 breadth. At one extreme is a leaf in which the length is only 

 one and two fifths times the breadth, making an almost circular 

 leaf; while at the other is a leaf whose length is three and one 

 half times the breadth, producing a very long narrow leaf By 

 dividing the leaves into groups based on the relation between 

 breadth and length, the following curve was prepared (Fig. i). 

 It should be stated that all the curves given here are based on 

 the same five hundred leaves. The figures along the base line 



* Illustrated with the aid of the Catherine McManes fund. 

 [No. 4, Vol. lo, of TcKKEYA, comprising pp. 77-100, was issued April 26, 1910.] 



101 



