64 THE INDIAN JOURNAL OF VETERINARY SCIENCE AND ANIMAL HUSBANDRY [ I, II

India, for a very large number of cattle here have been bred from the maimed, the
undersized, the unfit and the starved parents, and are bound to be unfit and useless
animals according to the basic law of breeding—" Like begets like ".

         Though the cattle of India measured by European standards are not economi-
cal or of first rate quality, they are without question the best tropical cattle in the
world. Their power to withstand epizootic diseases and ability to stand a hot
climate are factors of very great importance and value, and the best way of solving
the Indian cattle problem is to improve them without sacrificing these qualities. To
achieve this result it is necessary to establish and carefully maintain and breed herds
of the best Indian breeds and eliminate from them by a process of selection and
rejection all cattle below a certain fixed standard. This would gradually build up
profitable herds of indigenous cattle and is the direction in which the large land-
owners can best serve their country.

         The process of selection, however, is not as simple as it looks. Taking only one
character, milk yield, for example, there are very great variations possible
in it owing to the effect of environmental variations. For a proper selection it
would be necessary that the effect of these environmental variations on milk yield
be eliminated, or the milk yield records be first standardised to a definite common
basis. Whereas a good deal of statistical work has been done in other countries in
this line and results of practical application to cattle-breeders have been obtained
therefrom, no attempt has to the author's best knowledge been made in this country
so far to investigate how the milk yield of the best Indian breeds is affected by
these environmental variations.

         During the past 20 years a good deal of exact data regarding the milking capa-
bilities of hundreds of Indian dairy cows has been obtained on Military and other
Government farms. With a view to study the problem suggested above, a
statistical analysis of the vast amount of data available about Indian cattle
from these institutions offers to be a very fruitful line of research to a student
interested in cattle-breeding.

         Within the last few years Hammond and Sanders [1923] from Penrith Milk Re-
cording Society data and Sanders from Norfolk Milk Recording Society data [1927, 1 ;
1927, 2 ; 1928, 1 ; 1928, 2], have conducted statistical studies of complete lactation
milk records of English cows with the object of standardizing these for the four
factors, season of year, service period, age, and dry period which mainly affect them.
No attempt was made to study the effect of variations in management, it being
assumed that this would be minimised for cows of the same herd. The eliminating
corrections obtained by Sanders for Norfolk data [1928, 2] are given in Table I.