The writing of this little book has been delayed by the hope I once cherished of incorporating in it some of the results of the Indian Census of 1911. This desire was inevitable in the case of a retired Indian official, who, like most of his kind, has taken a small part in one or more of the decennial numberings of the Indian people. In this country, a Census affords material chiefly for the calculations and theories of the statistician, and the Registrar-General is not regarded as an expert in Anthropology or Linguistics. But in India the case is very different. If the district officer is always glad to learn as much as possible of the people with whom he is brought into contact, his official duties often reveal only the seamy side of Indian life, and it is only when he is in camp, or snatching a rare and hurried holiday in shooting, that he gets to see something of the people otherwise than as litigants or payers of revenue. A census is an agreeable and welcome opportunity for looking at India from another and more genially human point of view. In the first place, it is one of the least expensive of official operations, since it is chiefly performed by unpaid and volunteer agency. Hence the official, a little weary of litigants, touts, pleaders, and subordinates, who, however amiable in their private lives, are apt to be indolent and obstructive in office, is glad to make acquaintance with new friends, who, for the most part, take an intelligent and amused interest in the unfamiliar task of numbering. For many busy weeks before the actual counting takes place, the district officer has to ride far and near, to satisfy himself that all necessary preparations have duly been made, to issue the instructions that may be called for by the zeal, inquisitiveness or density of his volunteer colleagues. In the process, he has many pleasant and some amusing experiences. On one occasion I rode into a little village on the north-eastern frontier, inhabited by semi-savage Tibeto-Burmese people. Official orders as to the numbering of all the house in legible figures had apparently not been obeyed. I simulated wrath and disappointment, but the worthy headman on whom I vented my (purely official) indignation was not dismayed. "Bring out your drums!" he shouted. Every householder produced the family kettle-drum, on the head of which the number of his house had been duly inscribed in large figures. There was no paper in the village, but parchment was invented before paper, and the headman deserved the commendation I was glad to bestow. On another occasion, I found a house numbered indeed, but grievously dilapidated and obviously deserted. "Why is this empty house numbered?" I asked. "It is haunted by a ghost, sir," answered the enumerator. I confess I felt sorry not to allow him to include this ghostly visitant in a census of living men. Other incidents, more ethnologically important than these, will frequently occur. In any case the Census Report of an Indian province is by far the most interesting official document in existence, and each census adds something to our knowledge of Indian humanity, if only because each Census Commissioner, always an officer of unusual ability and attainments, looks at his task from a point of view somewhat different from that of his predecessors, and stamps his individuality on the work of his subordinates. Those who have read Mr E. A. Gait's article on Caste in the Dictionary of Ethics and Religion will expect the census of 1911 to contain new views and fresh information as to the actual working of the caste system in various provinces, and its relation to the religious ideas of the people.

It was natural, then, that I should wish to learn from a new tapping of the source from which has been compiled, for the most part, the ethnical portion of the first volume of the Imperial Gazetteer of India, which has been my chief authority in compiling this little book. But I know not when Mr Gait's Report for all India will be ready, and even the Provincial Reports come but slowly from the Press. Most of them are full of the most interesting and valuable information, but it takes time to assimilate so much new matter, and, in any case, not much of it could have been utilized for so small and elementary a book. Hence I have simply to state my debt to the late Sir H. H. Risley and Mr E. A. Gait for the chapter on Race and Caste; to Sir G. A. Grierson for the chapter on Languages, and to Mr William Crooke for enabling me further to summarise his masterly summary of what is known about Indian Religions. It is a particular pleasure to acknowledge my indebtedness to my friend Sir G. A. Grierson. Years ago, when we were young men, it was known that in him the Indian Civil Service possessed a scholar and a linguist of most unusual industry and ability. But few knew that there was germinating in his mind the scheme for the great Linguistic Survey of India, the most remarkable feat of administrative scholarship, perhaps, that has ever been attempted, a feat that has won him the Prix Volney and I know not what other appreciations of his work in France and Germany. His learning and linguistic skill are widely known, but I must seize the opportunity to tell of another feature of his achievement. Of course no man knows more than a few of the hundreds of Indian languages, but there is one man who knows something of the working and mechanism of them all, and that is Sir G. A. Grierson. I had the privilege of helping him with part of the Bodo volume of his Survey, having had occasion to learn one or two Tibeto-Burman languages in the course of official duty. The practised ease with which he acquired the syntactical and phonetic peculiarities of languages with which he had no previous acquaintance was the most surprising and delightful intellectual performance I have ever witnessed.

I have ventured occasionally to enliven my chiefly borrowed narrative with personal ideas or reminiscences. Such digressions have however been few and brief, and I do not think I need apologise for them.

I have to thank Miss Lilian Whitehouse and my son, Lieut. M. A. Anderson, R.E., for the two diagrammatic maps which will, I hope, clear up any geographical difficulties created by a necessarily brief account of a large and complicated subject.

I owe the illustrations of caste types to the kindness of Mr William Crooke. They are from photographs of inhabitants of one single district of the United Provinces and are interesting as showing how in a single small area racial differences show themselves in such a way as to be recognisable by the most careless observer. They prove once more how stratified Indian humanity has become under the influence of caste rules of marriages.