ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 827 



from the leaf-trace cambium, which again unite together the two 

 ruptured parts of the leaf-trace. This gives the appearance as if 

 the rupture took place on the upper side only of the leaf-trace, while 

 in fact it extends to the whole bundle formed during a period of 

 vegetation. The space left vacant by the rujjture of the leaf-trace 

 is probably filled up by the cambium, perhaps with the assistance of 

 the woody parenchyma wliicb surrounds the leaf-trace. The cells 

 thus formed, after their cell-walls have become thickened, constitute 

 the companion-cells (Berjleitzellen) which arise at a greater dej)th in 

 the wood. 



After the fall of the leaf the leaf-trace is completely ruptured, 

 exce2)t in the case of the Araucaricfe, where the leaf-trace does not 

 appear to be completely torn through, even in the oldest internodes 

 of the main stem. 



All the evergreen Dicotyledons examined, with the exception of 

 Aralia quinquefolia and Primus Laurocerasiis, have a thin-walled 

 tissue above the leaf-trace which separates it from the upper woody 

 portion of the branch or stem. This tissue is in immediate connection 

 with the pith, and consists of similar cells. Tlie leaf-trace is in all 

 cases curved outwards. In Ilex Aquifolium the leaf-trace is ruptured 

 in the third year. In this case the space left vacant by the rupture 

 is filled up by cells which resemble those of the pith-like tissue 

 lying above the leaf-trace. 



In all Dicotvledons which have a long portion of the leaf-trace 

 running through the cortex, a rupture of the cambium takes place 

 after the fall of the leaf. In Camellia Japonica the leaf- scar arises so 

 deep in the feebly developed cortex that there can hardly be said to 

 be a cortical portion of the leaf-trace. Hence the whole of it is in 

 this case concealed. 



The same is true of the leaf-trace in deciduous as in evergreen 

 Dicotyledons ; after the fall of the leaf a rupture of the leaf-trace 

 takes place at the cambium in those cases where it has a cortical 

 portion. 



Gum-canals of the Sterculiaceae.* — M. P. van Tieghem points 

 out that the more or less abundant production of gum or mucilage is 

 a common character of the Malvaceaj, Tiliaceaj, and Sterculiacese, 

 which ought, he considers, to bo treated rather as three tribes of one 

 order than as distinct orders. But the mode of formation of the gum 

 differs in the three groups. In the Malvaccje and Tiliaccse it is 

 Bccrotcd in large isolated cells which sometimes coalesce ; in the 

 Sterculiacea! in large schizogenous secreting canals. The cells 

 which border these ciinals do n(jt usually differ in any resi)ect from 

 the surrounding parenchyma, and may contain starch or calcium 

 oxiiluto ; sometimes tliey are smaller than the ordinary parenchy- 

 matous cells. These canals occur in the stem and leaves only, being 

 entirely absent from the root. In the stem they are usually developed 

 simultaneously in the cortex and in the pith ; the rest of the stem 

 being as a rule destitute of them. In the cortex they are arranged in 



* Bull. Soc. IVjt. France, xxxii. (1885; pp. 11-14. 



