for a number of years, that rules for the guidance of those concerned have been carefully framed, that checks are imposed on the subordinate officers responsible for the primary collection, and that there is no special reason why a subordinate officer should seek to exaggerate or reduce the numbers of livestock kept in his area. It is no doubt the case, as witnesses have asserted, that many enumerators, instead of actually counting cattle, resort to information collected from villagers, or to copying previous figures; but since the work is inspected and checked, this is to some extent a risky method of enumeration and it is unlikely that it is followed by a large percentage of the enumerators employed. While, therefore, it may be admitted that difficulties connected with the taking of a livestock census are considerable and that errors must be numerous, it should be pointed out that, in dealing with such very large numbers of village returns, errors due to exaggeration on the one hand and understatement on the other should tend to cancel out. It may also be claimed that the totals returned in recent years have been consistent. Where enumerators have failed, and to this failure many of the criticisms of their work must be due, is in the classification of the various groups of cattle. This is not altogether a matter for surprise, for, in asking subordinate officials to follow the classification prescribed, a difficult, indeed an impossible, task has sometimes been set them. In the Punjab, for example, where the village accountant is expected to compile his list of cattle under some twenty different categories, we find that, at the last census, 12,594 bulls were enumerated. The livestock experts of the province estimate the requirements at 50,000 bulls and, as there are nearly 2,800,000 cows, this seems a conservative demand. The figure for "bulls " in this province would be inexplicable were it not that, in another column, we find about 1,460,000 young uncastrated males recorded. It is evident that the Punjab village accountant has been set a perplexing problem by the authority responsible for the headings of the column in his village record. If we go to the other end of India, we find that the village accountant of Madras has been given a less invidious task than his fellow in the Punjab. His "return" of the cattle kept in his village contains eight columns only; but even he is asked to make a distinction which leads him into difficulties between "breeding bulls" and "other bulls." The Madras totals show that there are about 5,500,000 cows, but only 65,663 "breeding bulls." On the basis of the conservative estimate made by the Punjab livestock experts, it would seem that Madras should possess about 100,000 bulls fit for service. Animals under three years old are separately classed, so that the 2,134,000 "other bulls", which Madras village accountants distinguish from "breeding bulls," may contain from 30,000 to 50,000 animals which, in well regulated stock management, would be allowed to remain entire. In point of fact, in the conditions prevailing in Madras, many bulls are used for draught purposes and many of those enumerated among "other bulls" would subsequently be castrated and employed in the plough and cart; but the existence of over two million surplus bulls exceeding three years of age throws a significant light on the condition of animal husbandry in the presidency. DISTRIBUTION OF CLASSES OF CATTLE IN INDIAN PROVIN CES. 164. We have seen that the number of cattle kept in the different provinces, when related to the net area of land sown, shows a surprising variation. Before discussing the causes for this variation, it will be desirable to separate the total number of cattle into different categories. Since it is impracticable to distinguish " breeding bulls " from "other bulls " and "young uncastrated males," we class cattle of the ox tribe as " bullocks,' COWS and others." For buffaloes we have adopted the census headings in preparing the following figures: 99 66 Classes of cattle maintained in the major provinces of India and in British India compared with the net area sown in 1924-25 For British India as a whole, we estimate that for each 100 acres of net area sown there are 92 acres of uncultivated land available for grazing, to which should be added 21 acres of fallow and that on this total area of 213 acres there are supported 20 bullocks, 17 cows, 16 other cattle, 3 male buffaloes, 6 she-buffaloes and 5 young buffaloes, a total of 67 cattle, in addition to 27 sheep and goats and some other stock. Having regard to the very poor quality of the grazing available, and to the fact that it fails to afford adequate maintenance for cattle at the season of the year when fodder grown on cultivated land is scarcest, we are of opinion that this number of cattle is a heavy stock for land to for a number of years, that rules for the guidance of those concerned have been carefully framed, that checks are imposed on the subordinate officers responsible for the primary collection, and that there is no special reason why a subordinate officer should seek to exaggerate or reduce the numbers of livestock kept in his area. It is no doubt the case, as witnesses have asserted, that many enumerators, instead of actually counting cattle, resort to information collected from villagers, or to copying previous figures; but since the work is inspected and checked, this is to some extent a risky method of enumeration and it is unlikely that it is followed by a large percentage of the enumerators employed. While, therefore, it may be admitted that difficulties connected with the taking of a livestock census are considerable and that errors must be numerous, it should be pointed out that, in dealing with such very large numbers of village returns, errors due to exaggeration on the one hand and understatement on the other should tend to cancel out. It may also be claimed that the totals returned in recent years have been consistent. Where enumerators have failed, and to this failure many of the criticisms of their work must be due, is in the classification of the various groups of cattle. This is not altogether a matter for surprise, for, in asking subordinate officials to follow the classification prescribed, a difficult, indeed an impossible, task has sometimes been set them. In the Punjab, for example, where the village accountant is expected to compile his list of cattle under some twenty different categories, we find that, at the last census, 12,594 bulls were enumerated. The livestock experts of the province estimate the requirements at 50,000 bulls and, as there are nearly 2,800,000 cows, this seems a conservative demand. The figure for " bulls "in this province would be inexplicable were it not that, in another column, we find about 1,460,000 young uncastrated males recorded. It is evident that the Punjab village accountant has been set a perplexing problem by the authority responsible for the headings of the column in his village record. If we go to the other end of India, we find that the village accountant of Madras has been given a less invidious task than his fellow in the Punjab. His "return" of the cattle kept in his village contains eight columns only; but even he is asked to make a distinction which leads him into difficulties between "breeding bulls" and "other bulls." The Madras totals show that there are about 5,500,000 cows, but only 65,663 "breeding bulls." On the basis of the conservative estimate. made by the Punjab livestock experts, it would seem that Madras should possess about 100,000 bulls fit for service. Animals under three years. old are separately classed, so that the 2,134,000 "other bulls", which Madras village accountants distinguish from "breeding bulls," may contain from 30,000 to 50,000 animals which, in well regulated stock management, would be allowed to remain entire. In point of fact, in the conditions prevailing in Madras, many bulls are used for draught purposes and many of those enumerated among "other bulls" would subsequently be castrated and employed in the plough and cart; but the existence of over two million surplus bulls exceeding three years of age throws a significant light on the condition of animal husbandry in the presidency. 164. We have seen that the number of cattle kept in the different DISTRIBUTION OF CLASSES OF CATTLE provinces, when related to the net area of land sown, shows a surprising variation. Before discussing the causes for this variation, it will be desirable to separate the total number of cattle into different categories. Since it is impracticable to distinguish" breeding bulls " from "other bulls " and "young uncastrated males," we class cattle of the ox tribe as bullocks,' COWS and others." For buffaloes we have adopted the census headings in preparing the following figures: 66 99 66 66 Classes of cattle maintained in the major provinces of India and in British India compared with the net area sown in 1924-25 For British India as a whole, we estimate that for each 100 acres of net area sown there are 92 acres of uncultivated land available for grazing, to which should be added 21 acres of fallow and that on this total area of 213 acres there are supported 20 bullocks, 17 cows, 16 other cattle, 3 male buffaloes, 6 she-buffaloes and 5 young buffaloes, a total of 67 cattle, in addition to 27 sheep and goats and some other stock. Having regard to the very poor quality of the grazing available, and to the fact that it fails to afford adequate maintenance for cattle at the season of the year when fodder grown on cultivated land is scarcest, we are of opinion that this number of cattle is a heavy stock for land to carry. If the cattle are to yield a profit which would be accepted as satisfactory in countries where stock keeping is strictly economic, the bullocks would require to be fully employed, the cows to be of a heavy milking strain and the manure to be carefully conserved and returned to the land. The number of bullocks employed in different provinces may next be discussed. The number of bullocks actually employed on the land is, in all cases, less than is stated in the Table above, for the figures include bullocks utilised in towns, or by Government, or carting contractors. Although, in the aggregate, the number of these must be large, they can, in general, form but a small percentage of the total and their inclusion should not, therefore, cause any appreciable change in the relative position taken by the different provinces. The possible influence of large cities should, however, be borne in mind, in examining the figures. As is to be expected, the number of bullocks bears a close relationship to the number of cultivators and to the average size of holdings. provinces are arranged in order of the number of bullocks maintained per hundred acres, as is done in the following statement, it will be seen that the number of cultivators follows the same order fairly closely, and also that the number of cultivators who do not own bullocks must, in some provinces, be very large. If, again, the average area cultivated per yoke of oxen is compared with the average size of the holding, a close relationship is usually disclosed. The number of bullocks required in agriculture is also affected by the character of the soil, by the cropping adopted, by the length or shortness |