- 208 



internally up to a certain specific maximum varying for 

 different species, and that in this physiologically determined 

 limitation of salt storage they have a sufficient means for 

 securing the high osmotic pressure which FITTING has pointed 

 out in the desert-plants and which enables them to obtain 

 water from the soil. In addition to sodium chloride, there 

 is also excretion of carbonate of lime often in great quanti- 

 ties, so it is not at all certain that the excretion of sodium 

 chloride is of any special importance. And finally FITTING 

 points out that dew is exceedingly rare in the Sahara. 



The flowers of Reaumuria are large and terminal on the 

 branches, they appear in July. Below the calyx there are 

 several bracteoles forming an involucre. The fruit is a 

 capsule with about a score of white woolly seeds. 



The leaf anatomy of Reaumuria oxiana has been briefly 

 described by VESQUE. The thick epidermis consists of one 

 layer, and the stomata and salt-glands are depressed. The 

 leaf is isolateral with about two layers of palisade cells and 

 a central water-storing tissue. Between the green tissue and 

 the water-storing tissue, or amongst the palisade, are numer- 

 ous bands of sclerenchyma running longitudinally within the 

 leaf, and from these issue long, thin sclerenchyma-cells 

 which as idioblasts stretch through the palisade to the 

 epidermis as if to support it from the inside (fig. 46 A G; 

 also VESQUE tab. 8, fig. 7). 



Reaumuria fruticosa Bge. 



On moderately stable sandy soil I have found shrubs 

 of this species scarcely attaining the height of one metre. 



It is strongly branched, the branches being strikingly 

 thick and light in colour. The year-shoots may attain a 

 length of 5 centimetres, and apparently always die back at 

 the apex, which remains as a dry stick. Branches of two 

 different kinds occur on the year-shoot, some are short but 

 quite evidently branches with elongated internodes, others 

 are rosette-like short-shqots. The former may be wanting, 

 but when present they are generally placed towards the apex 

 of the year-shoot. They are rarely 2 centimetres long, gener- 



