9. The proportion of Musalmans to Hindus is as 1,000 to 587.8.
I have given the preceding table and its results because they exhibit the latest official returns of the population of the district; but I should add that the magistrate and collector expressed great doubt of the accuracy of the returns. The table contains internal evidence of error, of which the first series of figures relating to the thana of Bhawanigunge affords obvious examples. Thus in that police sub-division there are stated to be in all only 22,935 families, while the materials in men and women are at the same time said to exist of about 12,000 Hindu families and 38,000 Musalman families, in all 50,000 families—a difference which cannot be satisfactorily explained by supposing an unusually large number of widows and unmarried persons. Again, the Hindu men and women, are stated at about 12,000 each, and the Musalman men and women at about 38,000 each; on the other hand the Hindu children are made to amount to 86,000, giving about seven children to each Hindu couple, while the Musalman children are made to amount to only 33,000, giving less than one child to each Musalman couple—an excess in the former case, a deficiency in the latter, and a disproportion between the two classes which are irreconcileable with all experience and probability. In point of fact there were no checks whatever employed to guard against error, the magistrate requiring the return from the daroghas, and the daroghas from the zemindars; the zemindars employing their gomashtas or factors; and the gomashtas depending on the mondals or headmen and the chowkidars or watchmen of the villages for the desired information. Besides the unintentional errors that might be expected to arise in such a diluted process, executed in all its parts by ignorant and uninterested men, it is not improbably supposed that both landholders and cultivators are indisposed to make faithful returns whenever misrepresentation can escape detection. They have vague fears about the objects of such inquiries, the landholders apprehending an increase of assessment, the cultivators a requisition for their personal services, and both shrinking from that minute inspection of their condition which such inquiries involve. Without ample explanation, therefore, and without checks of any kind, it is vain to expect accuracy in such investigations.
While endeavouring to ascertain the amount of means employed for the instruction of the population of a given district, it is important to know how far those means come short of the object to be accomplished, i. e., come short of giving instruction to the whole teachable population. With a view to this result, one of my first objects was to ascertain the number of children between 14 and five years of age, which, after consideration and enquiry, I assumed to be the teachable or school-going age. It was evident that, having to deal in this matter for the most part with
