536 BOTANICAL GAZETTE [December 



woodlands. One of the most interesting features is a wood composed largely 

 of box (Buxus sempervirens) , possibly native to Britain. — H. C. Cowles. 



i 



Jurassic Osmundaceae. — Sinnott 12 has investigated the structure of 

 petrified stems of Osmundites Dunlopi from the Jurassic of New Zealand. 

 They are characterized by a parenchymatous pith, with no internal phloem or 

 endodermis, and leaf gaps are invariably present. In one specimen typical 

 diarch roots, with a definite cortex, were observed in the pith, and it is con- 

 cluded that the xylem elements described as forming part of a "mixed pith" 

 in O. Kolbei probably represent root bundles. It is thought that the relation- 

 ship between Zygopterideae and Osmundaceae cannot be close, since the 

 anatomy of the leaf trace and foliar bundle in the two groups is so widely 

 different. It seems evident that there were both pro tost elic and siphonostelic 

 Osmundaceae during the Mesozoic, and there is no evidence that the proto- 

 stelic forms have given rise to the siphonostelic ones. — J. M. C. 



Welwitschia. — Church 1 * has made a detailed study of the strobili of 

 Welwitschia, supplying facts and new illustrations that were much needed. 



i 



The "flowers" having clearly been originally bisporangiate, much attention 

 is given to reduction phenomena, such as dicliny, the diminished number of 

 members, etc. A good deal is made of "minimum construction" associated 

 with a xerophytic type, observed in the perianth, the androecium, and gynoe- 

 cium; but the statement is emphasized that "a minimum construction does 

 not imply any phylogenetic relationship to other minimum types." The 

 author finds no indication of any relation to the flowers of angiosperms. 



J. M. C. 



Relationships of Fagaceae. — Miss Berridge 1 ^ has investigated the flowers 

 of certain Fagaceae, especially those of Castanopsis chrysophylla, comparing 

 them with those of Castanea vulgaris, Fagus sylvatica, Quercus Robur, and Jug- 

 lans regia. She takes issue with the common conclusion that the Amentiferae 

 are an isolated group, with no obvious relationship to other angiosperms. She 

 shows that the flower differs in no essential from other epigynous types of 

 flowers, and compares in detail the flowers of Rosaceae and Cupuliferae, claim- 

 ing that in all probability the epigynous Rosaceae are most nearly related to 

 the ancestors of the Fagaceae. — J. M. C. 



W 



Ann. 



Botany 



x * Church, A. H., On the floral mechanism of Welwitschia mirabilis (Hooker). 

 Phil. Trans. Roy. Soc. London B 205: 115-151. pis. 9-13. 1914. 



f * Berridge, E. M., The structure of the flower of Fagaceae, and its bearing on 

 the affinities of the group. Ann. Botany 28:509-526. figs. 9. 1914. 



