ZOOLOGY AND BOTANY, MICROSCOPY, ETC 591 



separate, each set from the other. Melanorrhea has a leaf with scleren- 

 ehymatous hypoderm, 1-layered ; the vascular bundles are joined by 

 bands of sclerenchyma ; and there are numerous glandular hairs. Gluta 

 has a 2-layered leaf hypoderm ; bundles isolated ; and glandular hairs 

 are uncommon. 



The two last-named genera seem to approach Mangifera in their 

 affinities, through the species M.fmtida, M. andamica, and M. quadriflda, 

 in which a single-layered sclerenchymatous hypoderm is present. M. 

 Reba, with secretory canals in petiole and principal leaf-vein, indicates 

 connexion with Anacardium, as also do M. csesia and M. lagenifera: these 

 two latter, moreover, with sclerotic elements in the leaf-mesophyll, 

 approach Bouea, and this is substantiated by the isolated condition of the 

 vascular bundles in these, two species. Mangifera seems, therefore, to 

 represent a central synthetic type, combining in its several species the 

 anatomical characters met with in the other genera. 



Branch Types and the Age of Trees.* — M. Jean Daniel, in a 

 note, points out that certain trees bear three kinds of branch. 1. Long 

 branches, with elongated annual shoots separated by bud-scars arranged 

 at relatively long intervals. 2. Short branches, with dwarfed annual 

 shoots, the separating bud-scars being more or less crowded. 3. Mixed 

 branches, composed of an irregular succession of long and short shoots. 

 Two general types emerge, the Beech type and the Ginkgo type. The 

 author concludes from his study of these that the short branch differs 

 from the long branch in having a relatively small amount of wood, 

 in the absence of annual rings, in the fewness of medullarv rays, and 

 in the abundance of parenchymatous tissue. The mixed branch has 

 similar characters, but the difference from the long branch is less 

 accentuated. When the tree has attained its maximum growth, it pro- 

 duces only short and mixed branches, and consequently the annual 

 growth-increase is not clearly defined from that time forward. 



These considerations point to the impossibility of determining the 

 age of a plant with any approach to accuracy by observing the number 

 of " annual " rings. For more than one ring may be formed in a year 

 if the conditions are changed in certain ways ; where the rings seem 

 cpiite clear, again, their formation may last only for a limited time, and 

 in some cases the rings are very indistinct. 



Physiology. 

 Nutrition and Growth. 



Function of the Endodermis and Selective Absorption by 

 Protoplasm. f — J. de Rufz de Lavison publishes some interesting 

 experiments and conclusions relative to the details of root-absorption of 

 certain salts. The substances used were sulphate of iron, which cannot 

 pass through living protoplasm, and sulphocyanide of ammonium, which 

 can. Some striking drawings of magnified root-sections are given ; and 

 it seems clear that the suberized frame-work round the typical endo- 

 dermal cell is impermeable, so that every salt-solution which enters the 



* Cornptes Rendus, cl. (1910) pp. 1611-13. 



t Rev. Gen. Bot. xxii. (1910) pp. 225-41 (figs.). 



