218 



THE CO:\irOSITION OF SOME INDIGENOUS GRASSES. 



Alexander J. Taylor, B.A., A.I.C., 



Assistant Chemist, School of Agriculture, Cedara, Natal. 



Read July 11, 1922. 



Introduction. 



A considerable number of analyses of South African grasses 

 have already been recorded, but so far there has been but little 

 systematic work done in this direction. Marchand* * deplores 

 the lack of investigation into our native fodder plants, remarking 

 that " in most cases the composition of our native plants is 

 represented by the analysis of a single sample." There are so 

 many factors- -state of maturity, habitat, weather conditions — 

 that affect the composition and feeding value that a single 

 analysis must be regarded merely as pioneer work, as an indica- 

 tion of the approximate composition of the plant, although, of 

 course, the single analysis is true for that pai-ticular sample. 



Dr. Bews,^ in his " Grasses and Grasslands of South 

 Afiica," remarks: " Chemical analyses, therefore, to be of value, 

 should be repeated at intervals throvighout the season, and this. 

 so far as I am aware, has seldom been done." 



Some of the earliest investigations into the composition of 

 grasses of South Africa are those of Juritz^ and Ingle,'' " 

 which have been continued by the former, and by Vipond*^ 

 and Marchand.' Sawer^ in the Cedara Memoirs gives some 

 " Notes on the Botanical and Chemical Composition of the Veld 

 Herbage of Natal," dealing mainly with the veld of Cedara and 

 its neighbourhood. Here again the composition of the grasses 

 noted is based on a single analysis only. This work A\'as con- 

 tinued to some extent in later years, but no results were published. 

 During the period of the war the investigation lapsed until it was 

 again taken up by the writer in the Cedara laboratorv eai'K' in 

 1920. 



Collection of Samples. 



Samples of the grasses most commonly occuring on the veld 

 of Cedara Experiment Farm were collected during the year.? 

 1920-1922 at approximately the same period of each year, and 

 from areas of similar soil type, except, of coui'se, in the case of 

 vlei species. At the outset some difficulty was experienced in 

 obtaining samples at a stage of growth as nearly as possible the 

 same. Ultimately it was decided to collect early in the flowering 

 period, whilst anthers were showing, but before the seed had 



* These numbers refer to the papers listed in the References. 



