But it is still more curious to observe how men of intelligence and observation can be led, by the force of inherited opinion, into statements as to the effects of caste which are actually contradicted by their own experience. And in Mr. Raikes's interesting work, "Notes of the North-Western Provinces," we find an instance of how people will always attribute everything to this universal bugbear. Observing on the pride of high caste, "which withers whatever it touches," Mr. Raikes informs us that the Brahmins and Rajpoots of the rich province of Benares will not touch the plough owing to pride of caste. He next tells us that caste is little regarded to the north of Allahabad, where, from various causes, the demand for labour is greater. All of which, being traced to its true cause, simply amounts to this, namely, that where landed proprietors of good family are well off they naturally do not care to work, whereas in another part of the country where they are not well off, or cannot procure labourers, they do work. In the same way, the author, after telling us that infanticide has at one time or other been common all over the world, tells us that in India it is entirely caused by caste. Now, if we take caste to mean family pride solely, it certainly has influenced the matter, or at least tended to maintain the evil complained of; but I know of one instance, at least, in India where infanticide can be traced to satisfactory causes, and amongst a people who have always been observed to be remarkably free from what are called caste prejudices. The Toda tribe, on the Nilgiri Hills, are polyandrists, and, in order to keep down the number of the tribe, they naturally had recourse to female infanticide. This they have now abandoned, and my Toda guide very soon told me the reason. He said, "Formerly we used to kill the females, because we had little more than the produce of our buffaloes to depend on; but now that more people have flocked to the hills we can let our lands and get plenty to eat." He added, also, that the Government had ordered them not to kill their children; but, unless their means had improved, it is plain that a Government order would have had little effect. But, as regards this subject of infanticide, it seems to be a thing difficult to avoid, whenever conditions arise which are favourable to its extension; nor will repressive measures alone ever place any very complete check upon it. Like every other demand, it rises and falls with the necessities of the situation, and can never be originally caused by anything in the shape of caste feelings or regulations; and amongst these necessities I, of course, include the desire to avoid shame, or the prospect of shame in the family, or starvation, as well as the fact that women are an encumbrance to some tribes. Some people, I may add, are under the impression that polyandric habits, when once established, become necessarily a cause of infanticide.