286 VELD-BURNING EXPERIMENTS AT GROENKLOOF, PRETORIA. 



i.e., when the leaves first appeared above the .g^round, and, as a 

 knowledge of the succession of the species after a fire might 

 liave an important bearing on the problems of the change in 

 the veld due to burning, I had to overcome this difficulty. The 

 method adopted was to mark as many plants as p'ossible as they 

 appeared above the ground with a zinc label, about 6 inches 

 long, driven into the ground at the side of the shoot. Each 

 label was stamped with a consecutive number, and a book with 

 corresponding numbers, and with sufficient space between the 

 numbers to add notes, was kept for the records. As soon as 

 the plants flowered they were identified, and in this way the 

 rotation in which the various species appeared was known. This 

 method also proved useful in enabling one to study more or 

 less the jife-history of the species. It is true the objection 

 might be raised that it would be impossible to find all the labels 

 again, but in practice I found that if the labels were used freely, 

 the possibility of losing sight of any particular species during 

 the whole period of observation was done away with. This 

 method has also the advantage that the labels are permanent, 

 so that when the plots are periodically burnt, they remain in the 

 ground, and in this way one will be able to observe whether 

 any species disappears or increases through constant burning. 

 The plots were visited almost regularly every Sunday, 

 except during the month of October and part of November, 

 when the prevailing influenza epidemic prevented my going out. 



Observations, Plots E^ and Eo. 



Growth had already commenced on August 4th. and the 

 plots were covered with young green shoots just appearing 

 above the ground, but on Eg (the earlier burnt portion) growth 

 was more advanced than on Ei. The only plants in flower on 

 Ej were Lasiosiphon linifolius, Gerhcra piloscUoides, and 

 Gai^a-nia krebsiana, but while these species were also noted on 

 El, none had yet flowered. On both plots many grasses had 

 begun to shoot. A week later (August nth) Gaaania krebsiana 

 and Lasiosiphon linifolius were common in flcwer, especially at 

 the lower end of the plot, where the soil is a rich red loam. 

 Gerbcra piloselloides was seen occasionally in flower as well 

 as Morcca?,^. On E^ only a single specimen of Gacania krebsiana 

 was seen in flower. The following week (August i8th) there 

 was a marked change on both the plots E^ and Eo, and this was 

 more apparent on E^. where many of the species common on 

 Eo were beginning to appear. E^ really carried the same species, 

 but in a less advanced stage of development. On Eg, Gasania 

 krebsiana and Lasiosiphon (inifolius still retained the position 

 as the species flowering most freely, and were new fol- 

 lowed in this respect by Gerbera piloselloides. The following 

 additional species were taken in flower on E, : — Becium obo- 

 vatum,, Felicia muricata, fndigofera hcdyantha\ Tulbaghia 

 acutiloba. 



