226 Cd.Ml'OSITIOX OF SOME IN'DIGENOl'S GHASSES. 



Individual Variations. 



Species that were collected in successive years at approxi- 

 mately the same period of the year generally show a considerable 

 variation quite apart from the seasonal variation mentioned in 

 the second part of this investigation. In a few cases this was 

 so great as to exceed the difference between these species and 

 those of a similar composition. As mentioned at the outset, this 

 difference is to be expected, and only the repetition of analyses 

 can ensure a fair average. So far as the present data goes, it is 

 difficult to ascribe it to the climatic differences of the years in 

 question, as this variation in composition is decidedly irregular. 

 No doubt the ever-present difficulty of obtaining a fair average 

 sample accounts for part of the variation. 



Take the case of Digitaria sanguinalis, for example (see Table 

 IV). Here we find that the two samples taken in 1920 and 1921 

 are very similar in crude fat, fibre, protein and carbohj^drates, 

 although the earlier sample is slightly poorer in each of the 

 valuable constituents and slightly higher in fibre, this cumulative 

 deficiency being reflected in the calorific value. The correspond- 

 ing sample taken in 1922 (that of 7/3/22) shows better results 

 throughout, but on the whole the three analyses are fairly con- 

 cordant. Also in 1921, within four days of collecting the sample 

 mentioned above, some Digitaria sanguinalis was found growing 

 on a neglected piece of ground that had been heavily mamu'ed, 

 where it had become most luxuriant in growth. The composition 

 of this sample was found to differ so markedly from all other 

 samples of the species that it has not been included in the 

 average, but shown separately in Table I. Here every con- 

 stituent — except crude fibre and soluble carbohydrates — shows 

 lower percentages than do any of the other samples of the species. 

 Of counse, the total feeding value |)er acre would greatly exceed 

 the average, but that side of the question has not been dealt with 

 by the writer. 



In Panicuni lacvifolium greater variation exists — for example, 

 crude fibre amounts to 39'1 per cent., 31'7 per cent., and 36"7 

 per cent, respectively in the three years. Protein shows a 

 maximum in the second year and a minimum in the first, 

 although the third year comes very near to the composition of 

 the first The marked general inferiority of the first sample is 

 reflected in the calorific value of 2,250, compared with the 

 average of 2,436 for the whole species. 



Setaria imberhis shows a decidedly superior composition in 

 1921 as compared with the following year, the latter showing a 

 marked falling off in protein especially. As in the case of D. 

 sanguinalis, a very luxuriant sample was collected in 1921 — on 

 the same day and near the same area as the sample of average 

 growth. The composition of this sample, however, is not so 

 abnormal as in the case of D. sa)igiiinalis. It shows a slight 

 inferiority to the average of that year, but not to such an extent 

 as does the sample collected at the same period in the following 

 vear. 



