260 THE INDIAN JOURNAL OF VETERINARY SCIENCE AND ANIMAL HUSBANDRY [ IV, III

(c) Fibre-length. In the determination of mean fibre-length, the method of
zoning and measuring elaborated by Roberts [ 11 ] was used. The original
samples were each divided into eight zones, and the proportions of fibres measured
from the final sub-samples were 1 in 3 or 1 in 4, in the pre-experimental samples,
and 1 in 8 or 1 in 10 in the samples taken at the end of the experimental
period. The fibres were measured to the nearest half-centimetre.

In zoning and in counting out the fibres to be measured, great care was exer-
cised to reduce as much as possible the chance of breaking fibres. In some
samples, especially among those of Group I, taken after the feeding period,
a considerable number of very short fibres (of less than half the mean
fibre-length) were found. These caused the distribution of fibre-length in the
particular samples to exhibit marked skewness, but their contribution to the total
lengths of the fibre was small. (The greatest proportion was found in animal
number 4, Group II, after experiment, where short fibres of 2 cm. or less, formed
34.9 per cent. of the fibres measured ; their contribution to the total length was,
however, only 6 per cent.) In measuring, the short fibres were treated as
separate groups, and they are not included in the figures for mean fibre-length,
shown in Table V. Since the total length enters into the determination of fineness,
the short fibres are involved in the calculation of the results shown in the succeed-
ing sections.

                                             TABLE V.

              Mean fibre-lenglhs (in cm.) excluding short fibres.

Differences of means ; pre-experimental, 0.33 ±0.47 ; experimental, 1.02 ±0.30 ; percentage changes, 8 .4 ±3.56.