May, 1904.] Deciduous Leaves. 165 



thus have a longer time to be withdrawn into the stem. In the 

 Catalpa, for instance, the blades often freeze and dry up in the 

 fall while the petioles are still green and active. It would evi- 

 dently be better if the useless blade were cut off by a cleavage 

 plane so as not to hinder the work of the petiole. 



In compound leaves the leaflets are usually- shed singly. The 

 leaflets of such palmate leaves as in the Virginia Creeper and the 

 various Buckeyes are cut off some time before the petiole. Pin- 

 nately compound leaves ha\ e various peculiarities. In such forms 

 as Rhus glabra the leaflets are separated by cleavage planes but 

 no transverse cleavage joints are formed in the main rachis which 

 persists for some time. In other forms, like in Fraxinus quad- 

 rangulata and Staphylea, not only are the leaflets cut off by 

 cleavage planes but there is a series of cleavage joints formed in 

 the rachis at the insertion of each pair of leaflets and thus the 

 main rachis of the leaf drops off piece by- piece. Decompound 

 leaves often form an elaborate sy'stem of separation layers. A 

 good example of this is shown in the leaf of the Honey L,ocust. 

 First the numerous leaflets drop off, the main rachis and the side 

 branchlets remaining on the tree for some time. Next the side 

 branchlets begin to fall, and finally the whole rachis is separated. 

 One may well a.sk the meaning of such an elaborate system of 

 cleavage planes when one amputation at the base of the petiole 

 would be sufficient. There is no doubt but that the green rachis 

 and petiole may continue, to a limited extent at least, the process 

 of photosynthesis ; and as stated above, by means of a gradual 

 cutting away of the large leaf surface the more exposed parts are 

 removed first and there is a better opportunity for the withdrawal 

 of the food present into the stem. 



A very interesting condition is present in the Green Briers. 

 The leaf of Smilax hispida has two tendrils near the base of the 

 petiole and these, of course, hold the plant to its support. Evi- 

 dently if the leaf were shed in the usual way the whole vine 

 would fall to the gouna in the winter. There is a more or less 

 perfect brittle layer formed in the petiole just a little beyond the 

 two tendrils where the leaf finally breaks off, leaving the petiole 

 base with the tendrils intact. Most of the leaves hang on until 

 after December i, though usually frozen before this time. The 

 development of a brittle layer in the petiole of this plant seems to 

 be quite a modern adaptation. Smilax glauca, S. rotundafolia, 

 and S. bona-nox show the same peculiarity. The genus Rubus 

 represents another group of plants which .shed their leaves by a 

 break in the petiole, leaving the base on the stem. In this case 

 there are no tendrils and the only apparent advantage to the plant 

 is the protection of the bud or tender part in the axil of the leaf. 

 The adaptation, however, may have no other significance than one 

 of the po.ssible ways in which the plant was able to get rid of its 



