116 PROCEEDINGS OF THE ACADEMY OF 



in the Ranal alliance, strong as it always has been acknowledged to be, is 

 still more strengthened. 



It is impossible to suppose that a so closely allied genus as Liriodendron 

 should be founded on a different type from Magnolia. We shall see that only 

 very slight causes, which we can well understand, have made some of the 

 chief foliar distinctions, and the few which we cannot prove from actual facts, 

 can be made almost certainties from parallel observations. The identity of 

 type will in this way be manifest. 



First, as to the premorse or cut off appearance of the end of the leaf blade. 

 This all results from the stipular portions being adnate with the stem axis, 

 instead of being wholly on the petiole as in Magnolia. In the latter the stip- 

 ules are carried along as the petiole advances, the leaf blade cannot grow be- 

 yond, and so in vernation has to lie flat up against them. In Liriodendron, the 

 stipules being fast to the main stem, the petiole carries the leaf blade beyond 

 them, over which it is bent until its apex is brought down in contact with the 

 straight line formed by the union of stipule and stem. Here it is pressed as into 

 a mould by the elongating petiole, and the form of the leaf which we see is the 

 necessary result. These processes in Magnolia and Liriodendron can readily be 

 seen on an examination of the buds at any time during the growing season ; and 

 to those who have no specimens the figure of the latter in Gray's Genera will 

 easily give the idea. It may be here noted that those who look only to Mr. 

 Darwin's principle of natural selection to account for the laws of form, might 

 be troubled by such cases as these. It is scarcely conceivable that a square- 

 edged leaf blade, as we find it in Liriodendron, is of any special benefit to the 

 species ; yet if this form is the consequence of some other act, which is a ben- 

 efit, the selection principle may still hold. 



If the ternate type of leaf is probable in LAriodendron, as in Magnolia, the 

 lower portion of the petiole, and lateral or stipular portions, must have ad- 

 nated with the stem prior to the full development of the leaf. This view ne- 

 cessitates the idea that the leaf does not always originate at the node from 

 which it seems to spring. I do not believe it does; but I am well aware that 

 in this I have opposed to me the weight of our best botanical authorities, from 

 whom I would not yet dare to positively differ until I shall have the weight of 

 more facts. 1 would only say that in the case of Liriodendron the appearances 

 are much in favor of the belief that in an early stage the petiole clasped the stem, 

 and for a considerable length ultimately became an integral part of its cortical 

 system. The vessels which are seen connected in direct lines with the peti- 

 oles below and above the node, as they are in existence before the leaf bud 

 has opened, and the leaf blade has had any chance to elaborate sap from the 

 light or air, supposed to be necessary just above before they could be formed, 

 do not seem to originate at the node ; while the fact that these vessels suddenly 

 curve from the opposite side towards the supposed petiolar base is much more 

 characteristic of an unfolding sheath than of a descending current of matter 

 which would most naturally go down in astraightish line. But that the peti- 

 ole has really adnated with the stem in this way in Liriodendron seems most 

 probable from the fact that on the opposite side from the leaf is often seen a 

 ridge which could hardly be formed except by the meeting of two edges en- 

 closing a stem, with a little to spare ; and at other times there is a slight de- 

 pression, as if the two opposite edges barely met. There seems to be every 

 evidence short of an actual witnessing of the fact, that the petiole in Lirio- 

 dendron became adnate with the stem, and in this way the two lateral sections 

 (stipules) were brought in contact with the siem with which they united. This 

 would bring them nearer the sources of nutrition, and enable them to assume 

 ^ more leaf- like and permanent character than if on the petiole. They become 

 rather primary than secondary leaf organs, and this is just what we see them 

 to be. 



Thus we may assume that Magnolia has typically a ternate leaf structure ; 

 that the stipules are the two lateral lobes which by a peculiar process of ad- 



[Oct. 



