AN (ECOLOGICAL SURVEY OF THE MIDLANDS OF NATAL. 525 



do not penetrate to any great depth. This is due for the 

 most part to the fact that over most of the thorn veld the 

 hard shales come very near to the surface. Where the soil 

 happens to be deep they do penetrate to considerable depths. 

 The different species, however, vary in this respect. There is 

 also very considerable variation in the degree of spine- 

 development in the different species and in different indi- 

 viduals of the same species. The thorns vary in both size 

 and number. The more thorns the less leaves. There are 

 also indications of a correlation between the nature of the 

 root-system and the degree of spinosity, but the whole 

 subject is one to which the writer hopes to give further 

 attention. 



Another question of great interest is how far have the 

 thorn trees been influenced in their growth-forms, etc., by 

 the constantly recurring grass fires. It was pointed out that 

 grass-burning was the usual practice when Vasco de Gama 

 landed in 1497. It has certainly been going on regularly 

 ever since. The thick bark which undoubtedly does help to 

 protect the thorn trees from the fires, may be, at least, 

 partly a direct adaptation to environment. Their seeds are 

 very hard, and under ordinary conditions will lie for a long 

 time in the soil without germinating. After a grass fire, 

 however, the hard seed-coat is burst by the scorching and 

 they germinate immediately. The young trees grow quickly, 

 and a few of them manage to grow tall enough to be able to 

 resist the next grass fire. The bulk of them are killed, and 

 consequently we do not find them growing close enough to 

 form thorn-thickets. The soil is uniformly very dry, and their 

 growing isolated is not a result of any differences in soil 

 moisture. Altogether the thorn veld presents some of the 

 most interesting problems in connection Avith the plant 

 oecology of Natal. 



VOL. 2, PART 4. 35 



