308 Annals of the Carnegie Museum. 



most lobes are branches from the two lateral primary veins and are 

 therefore secondary veins. This branching of the primary veins is 

 different in the other allied genera of the Araliacccc, StcrculiacecB, 

 Menispermacecc, Vitacecu, and Haniamelidacecc {Liquidamhar). These 

 are all, with very few exceptions, if at all palmately veined, of the 

 palmately five-veined type. This is also evident in the leaves not 

 lobed. In the five-lobed leaves of these last-named genera the veins 

 running to the two lowermost lobes are primary veins, or, if they are 

 secondaries, branching from the two upper lateral primaries, the origin 

 is always very close to the midrib. 

 The data from my own observations are as follows : 



a. Stcrculiacccc. — The leaves of thirty-one species were examined. 

 In thirty of these the leaves were of the palmately five-veined form, 

 or showed a very strong tendency toward it, while only one species 

 showed three palmate veins, and this one showed no tendency to 

 lobing. 



b. Araliaccar. — Out of sixteen species examined, the leaves of four- 

 teen were of the palmately five-veined form ; one with three veins, but 

 no tendency toward lobing ; the other one, Panax trifolia, shows many 

 of the leaflets to be divided into four and five smaller leaflets, and in 

 each case they are palmately veined, even those leaflets with only 

 three smaller leaflets showing the five palmate veins. 



c. Vitaceaic. — Out of twenty-seven species examined, the leaves of 

 twenty-three species show the palmately five-veined form. A few of 

 these have a tendency to form seven lobes; in this case the veins 

 running to the two lowermost lobes are secondary veins. One species 

 showed three veins springing from the base of the leaf, but in this case 

 the leaf was a pinnately compound leaf. In one species (Cisstis 

 repcns) the two lowermost veins come from the two lateral primaries, 

 but originate very close to the midrib. In the remaining two species 

 there are three veins, with but little, if any, tendency for others to 

 branch from them; if so, they branch very close to the midrib. 



d. Menispermacecc. — In this family eleven species were examined, 

 and all leaves showed the five — or more — palmate arrangement of 

 the veins. 



e. Hamamelidacecc. — Sixteen Liquidambar leaves, all of one species, 

 were examined. Of these twelve were palmately five-veined and five- 

 lobed, three were palmately three-veined and three-lobed (with no 



