COMPOSITION OF SOME IXDIGfiXOUS GRASSES., 225 



On examining the relative values in protein content this 

 similarity of composition, whilst not so evident, may be still 

 observed. The highest crude protein content is shown by 

 Axonopus scmialatus with 9-8 per cent., and the lowest by 

 Andropogon amplectcns with 3'7 per cent., a wide variation, but 

 again omitting the extremes there are ten species from Panicum 

 ecklonii (9"2 per cent.) to Anthistiria imherbis (5"9 per cent.), 

 which lie fairly close together. Only one ruderal species falls 

 within this limit, and most of the pioneer species and post-climax 

 grasses lie below it. 



So far as the present data goes, one is led to the conclusion 

 that the different species of the climax veld type all tend to a 

 very similar composition. This is true also of the ruderal and 

 pioneer types, but to a decidedly lesser degree. Species of these 

 types, especially the former, have a more varying habitat than 

 in the case of the climax veld type, and so have greater scope 

 in which to assert their individuality. 



Transverse Sections of the Leaf. 



With regard to the method of ascertaining the relative values 

 of grasses and similar feeding stuffs, to place them in inverse 

 order of their fibre content appears at once to be simple and satis- 

 factory. This method gives results that are concordant with 

 other methods — e.g., calorific value — and with general agricultural 

 opinion based on practical experience. A study of Table II will 

 illustrate this point. 



Dr. Bews^ states that " much may be learned by studying simple trans- 

 verse sections of the leaves of the grasses. It is much less laborious and 

 probably even more useful on the whole than elaborate chemical analyses of 

 the herbage, for these, without doubt, vary greatly according to the time 

 of the year end even according to the state of the weather. ... A glance 

 (at the cross section) is sufficient to show whether eacli grass represented 

 is likely to prove palataljle to stock. The less sclerenchyma shown the 

 ni_ore useful is the grass from this standpoint." 



The sclerenchyina, of course, corresponds to the " crude 

 fibre " of the chemical analysis, and thus the analyses made 

 confinn this opinion. Bews figures cross sections of a number of 

 the species, or of closely allied species dealt with by the writer, 

 viz., Atidropogon hirtus, Anthistiria imberbis, Aristida bipartita, 

 Digitaria ternata (allied to D. horizontalis and D. sanguinalis), 

 Eragrostis curvula, Harpechloa capensis, Sporobolus indicus, 

 Tricholaena rosea, and Tristachya leucothrix. A classification of 

 these species based on the cross sections as figured would place 

 the species much in the same order as they are found under tlie 

 heading of crude fibre in Table II. 



This method does not taken into account the relative iDrotcin 

 content, which in some species — for example, Rottboellia com- 

 pressa, Eragrostis chJoromelas — differs widely- from the relative 

 fibre content. Thus a very important factor from the stock 

 farmer's standpoint is neglected. 



