THE PLANT SUCrKSSION IN lUK TUORX XELD. 1 6Q 



shoots, and doubtless sometimes are agents in distribution. 

 Sclater {loc. cit.) describes 36 species of South African ante- 

 lope, but many of them, though formerly abundant, are now- 

 rare or almost extinct. The African elephant, now in South 

 Africa confined to the Addo scrub, must formerly have been 

 an important agent. Its food consists of leaves and twigs, wild 

 fruits and roots, seldom grass. The Cape gerbille {Gerhillus 

 afer) is very common, and portions of the veld may be riddled 

 with its burrows. Its food consists of bulbs and seeds. The 

 rats, of which the commonest is apparently a brown variety of 

 the black rat (AIiis rattus) are omnivorous, and may be of im- 

 ]>ortance. Ants (true ants, not termites) are very abundant in 

 the Thorn Veld, and apart from their effect on the soil, some 

 of them store seeds which they may bring from a considerable 

 distance. Plants, the seeds of which have special adaptations 

 for dispersal by ants, such as a brightly-coloured caruncle con- 

 taining oily food-material (elaiosome). are known as myrme- 

 cochorous plants. Sernander * has written a monoiiraph of 

 European myrmecochorous plants, reviewed by Weiss in the 

 Nezv Pliytologisf, 1908. \'Veiss+ has also described the dispersal 

 of the seeds of the gorse and the broom by ants. I have made 

 a number of preliminary experiments by placing seeds of various 

 plants near a nest of the ant Tctramorimn squainiferum. In 

 the nests of this ant collections of seeds are found stored in 

 special chambers. Grass seeds predominate. When ten seeds 

 of the grass Paspalnm scrohiculatum were placed near a nest, 

 they were all removed in about nine minutes. Other grass seeds 

 were also readily carried away. The order of ijreference was 

 then determined by placing eight different kinds of seeds (six 

 of each), and observing how many were left at intervals of 15 

 minutes. The experiment was repeated several times and at 

 diff'erent nests. Other experiments with different sets of seeds 

 were also carried out. The following seeds were observed to 

 be carried away readily: Paspaluw scrohiculatum, Phalaris 

 aruudiuacca. Hibiscus trionum, Teucrium riparium, Abutilou 

 sonncritianuni, Argemonc mexicarm, Sida rhoynbifolia, Nicandra 

 physaloidcs. hidigofcra sp.. Datura stratnomuni. Sonchus sp. 

 K\en fairly large seeds like Cassia occidentalis were taken, 

 sometimes two ants co-operating to remove one seed. Small 

 seeds, on the other hand — e.g., Amarantus retroflexus, Chenopo- 

 dium botrys, Plantago major, Rumex obtusifolius — were not 

 touched. On examining those various seeds under the micro- 

 scope, in the majority no special elaiosomes or other adaptation 

 could be distinguished. However, in Hibiscus trionum there 

 are little tufts of hairs, some of them glandular, dotted over 

 the surface of the seed. In Teucrium. the whole surface of the 

 seed is covered with glandular hairs, which microchemical tests 



* Sernander, R , " Entwurf einer monographic der Europaische 

 myrmecochoren." KuiisL S'zynska Vetenskapsakademicns Handlingar 

 <i9o6). 



t Weiss, F. E.. New Phyfologist. 8, f,^], (rgog). 



B 



