K.— BOTANY. 197 



characters. Several such physiological strains are now known in South 

 African species of the genera Pentzia and Salsola. There are two 

 strains (physiological varieties) of Salsola glabrescens Burtt Davy, which 

 grow side by side. One of these plants, with purplish-red young twigs, is 

 closely grazed, while the other, in which the young twigs always appear 

 to be pale-coloured, remains untouched by cattle or sheep until there is 

 nothing else to eat. The selective feeding of the animals when grazing 

 on a pasture bearing these two forms is very remarkable. Carefully 

 collected dried specimens of these two plants, with flowers and fruit, have 

 been critically examined at Kew, but apart from the colour of the young 

 stems of the living plants, no single character can be found by which one 

 form can satisfactorily be separated from the other. In the fresh condition 

 the only distinction between typical specimens of the two forms mentioned, 

 and also in two forms of another species of Salsola, is that of the stem- 

 colour, and neither form possesses any odour that can be detected. It 

 would be of great interest, therefore, could we discover how the animals 

 are able to distinguish the palatable from the unpalatable form, since we 

 might then become as acute as they appear to be in appreciating the 

 significance of fine distinctions. 



Certain South African species of the genus Pentzia afford another 

 interesting case of the superior discriminating power of animals over 

 botanists. 



The Pentzias in question are strongly scented. It was noticed that 

 sheep and caterpillars were feeding on a large stretch of country covered 

 by these plants, and that some of them were being eaten while others 

 were avoided. Representative specimens of all those that were being 

 eaten and of those that were avoided were collected by our present Assis- 

 tant for South Africa and brought to Kew for critical examination. No 

 specific difference between the different ' types ' could be detected during 

 a preliminary examination in the field, though they could be recognised 

 one from another by their external appearance. 



Detailed field notes of each specimen were made on the spot for use in 

 the herbarium investigation, since it was found that some of the plants 

 were greedily devoured, some were most carefully avoided, while others 

 were usually left untouched ; when, however, the latter were grazed, as 

 sometimes happened, unmistakable symptoms of nervous depression were 

 produced in the animals. 



On Taxonomic grounds we must regard all three forms as being 

 members of one and the same species, since no morphological difference of 

 any value can be detected between them. 



Then again there are puzzling problems connected with the character 

 of certain species on different types of soil, for it has been noticed in South 

 Africa that, while a species may be a useful pasture plant on, say, a red 

 loamy soil, yet when the same species, growing on tufaceous limestone, is 

 eaten by stock a heavy mortality may result. Here, then, is another 

 interesting problem in the domain of applied botany and soil chemistry, 

 which, like the cases mentioned earlier, may also fall into the domain of 

 Taxonomic Botany. 



It is also very remarkable that the Indian Lac insect {Coccus lacca) 

 has drawn our attention to the existence of two physiological forms of 



