COMPOSITION OF SOME INDIGENOUS GRASSES. 281 



uted much of the cattle diseases of this country — it is of interest 

 and value to obtain definite figures as to the amount of this 

 important constituent of the ash of the common veld grasses. 



Unfortunately, as this phase was only taken up this year, 

 results are based on a single sample only, except for those results 

 appearing in Table IV. In the ordinary climax grasses the 

 amount of phosphate present is remarkably constant, averaging 

 a little over 0-2 per cent, of the absolutely dry grass. In the 

 ruderal species the figures show considerable variation, averaging 

 between 0'3 per cent, and 0"4 per cent. 



By way of comparison the following figures are quoted after 

 recalculating to dry matter: — 



Meadow hay (Crowther'"), 0'46 per cent, phosphoric oxide. 

 Pasture grass (Crowtlier''^), 0-60 per cent, phosphoric oxide. 

 Teff hay (Vipond*^), 0-22 per cent, phosphoric oxide. 



The method of determination adopted was Vipond's*^ modi- 

 fication of the methods of von Lorenz and Pemberton. The ash 

 as obtained in the general analysis was taken up with dilute nitric 

 acid, the phosphate precipitated as phosphomolybdate, dissolved 

 ui a known volume of N 10 KHO solution, and the excess titrated 

 back with N/10 H.SO.i solution, from which the amount of 

 phosphoric oxide is calculated. 



Summary. 



1. Piuderal grasses usually have a better composition and a 

 higher feeding value than do the ordinary grasses of the 

 climax veld, whilst pioneer and post-climax grasses are 

 generally coarse and inferior in feeding value. 



2. The composition of grasses of the climax veld tends to 

 uniformity, that of the other types, esi^ecially ruderal, being 

 more variable. 



8. Individual and seasonal variations are considerable, especially 

 as regards the ruderal species. These variations are iiTegular 

 and cannot be accurately ascribed to any definite cause or 

 causes until further data is available. No doubt the ever- 

 present difficulties of sampling may account for a part of 

 these variations. 



4. The above facts may be explained and correlated to a con- 

 sideraible extent if they are viewed in the light of the differ- 

 ences of the species in growth form and in ecological life 

 history. 



5. Antliistiria imhcrhis (Thcnicda friandni), the principal veld 

 grass from the stock farmer's point of view, has also about 

 the best composition and feeding value of the grasses of its 

 type. 



6. The crude fibre content affords probably the best criterion 

 for determining the relative feeding values of grasses, the 

 feeding value being in inverse ratio to the fibre content. 



