This evidence, however, needs corroboration. Meanwhile, considering how dangerous it is to generalise on the strength of stray statements and observations of foreigners, considering how these statements reveal only half-truths in the case of many Indian customs, we may take this evidence with caution. If, however, what Hannay says is based on the actual practices and beliefs of the Persian Mahomedans, we are led to infer that not only in several places in Europe but also in many parts of Asia the rag came to be regarded as a vehicle of disease, whatever may have been the original ideas underlying the offering. When we have evidence to show that in Europe pins and rags were used at wells for purposes of divination, it is not difficult to conceive the process of reasoning by which these articles came to be regarded as appropriate offerings to the indwelling spirits, no matter how insignificant their intrinsic value. These instruments of divination, having done their duty, must have been consigned to the waters as being the best place for depositing them. Then, probably, they were looked upon as indispensable offerings to the water goblins and then, although the practice of divination disappeared, these articles still came to be regarded as appropriate offerings for the well-spirits, and the rustic mind, ignorant of the genealogy of the custom, interpreted the survival of the ancient usages according to its own conception of sympathetic magic and either looked upon the rag-offering as a charm for disease-transference, or as a connecting link with the deity. This theory of the origin of the custom, which is here put forward with some diffidence, also explains the growth of the two conflicting theories (1) that the rags are vehicles of disease or charms for disease-transference, and (2) that they are simple offerings to propitiate the deity. We have rambled far in our survey of the cult of rag offerings, because it represents a peculiarly interesting phase of water-worship. Of the rituals practised in the worship of water divinities it is the most rude and primitive. While the ceremonies of well-worship in the west correspond in several details, notably the offerings of flowers and coins, to the rituals with which we are familiar in this country, they contain some distinctive elements, the most remarkable of which is this practice of rag-offerings. We have cited numerous instances to show how common the practice was in Great Britain and Ireland and how it survives even to this day in certain parts, but while it was and still is a general feature of water-worship in those parts, it was and is unknown in India, although some folklorists have erroneously identified it with the entirely different, though seemingly analogous, practice of hoisting flags or _dhajas_ at shrines and sacred trees. Perhaps the best explanation for the practice of rag-offering is that it may be a degenerate form of flag-offering. CHAPTER XIX. ANIMAL DEITIES OF WATER.