The Todas

by W.H.R.Rivers

The custom of infant marriage is well established among the Todas, and a child is often married when only two or three years of age. When a man wishes to arrange a marriage for his son, he chooses a suitable girl, who should be, and very, often is, the matchuni of the boy, the daughter of his mother's brother or of his father's sister. The father visits, the parents of the girl, and if the marriage is satisfactorily arranged he returns home after staying for the night at the village. A few days later the father takes the boy to the home of his intended wife. They take with them the loin-cloth called tadrp as a wedding gift and the boy performs the kahnelpudithi, salutation to the father and mother of the girl, and also to her brothers, both older and younger than himself, and then gives the tadrp to the girl. Father and son stay for one night at the girl's village and return home on the following morning. Sometimes the girl returns with them to the village of her future husband, but, much more commonly, she remains at her own home till she is fifteen or sixteen years of age.

If a man has not been married in childhood he may undertake the arrangement of his marriage himself, and visit the parents of the girl unaccompanied by his father; and in this case the girl may at once join her husband if she is old enough.

From the time of the child-marriage the boy has to give a tadrp twice a year until the girl is ten years old, when its, by a putkuli. The tadrp which is given at first is very small, worth perhaps only four annas, but as the girl becomes older it is expected that the garment shall become larger and more valuable.

If any member of the girl's family should die it is expected that the boy's family shall on each occasion give a sum of eight annas or a rupee. This gift is called tinkanik panni litpimi, or " we give a piece of money to the purse."

Formerly the boy's family had also to contribute one of the buffaloes killed at the funeral, but this custom is now obsolete. The contribution of buffaloes and money from the boy to his parents-in-law is called podri. The boy has to take part in a ceremony at the funeral in which a cloth is laid on the dead body, and with this ceremony there is associated a further gift of one rupee, paid to the relatives of the dead person by the family of the boy who has married into the family of the deceased (see P. 358).

Certain ceremonies are performed shortly before the girl reaches the age of puberty. One is called puttkuli tazar utiti, or "mantle over he puts," in which a man belonging to the Tartharol if the girl is Teivali, and to the Teivaliol if she is Tarthar, comes in the day-time to the village of the girl and lying down beside her puts his mantle over her so that it covers both and remains there for a few minutes.

Fourteen or fifteen days later a man of strong physique, who may belong to either division and to any clan, except that of the girl, comes and stays in the village for one night and has intercourse with the girl. This must take place before puberty, and it seemed that there were few things regarded as more disgraceful than that this ceremony should be delayed till after this period. It might be a subject of reproach and abuse for the remainder of the woman's life and it is even said that men might refuse to marry her if this ceremony had not been performed at the proper time.

It is usually some years later, when the girl is about fifteen or sixteen, that she joins her husband and goes to live with him at his village. The parents of the husband announce that they will fetch the girl on a certain day, which must be one of two or three days of the week,' different for each clan. The husband, accompanied by his father and a male relative of the same clan, goes to the village of the girl, and the three are feasted with rice and jaggery. The husband puts five rupees into the pocket of the girl's mantle and then takes her home. There is no ceremony of any kind, not even the salutation such as was performed at the original ceremony.

If the youth does not wish to live with the girl when the time arrives, he may annul the marriage by giving one buffalo as a fine (kwadi-) to the girl's parents ; but, on the other hand, the parents of the girl have to return as many buffaloes as he may have given as pidri at funeral ceremonies.

If the girl refuses to join her husband the fine is heavier, and at the present time usually amounts to five or ten buffaloes, the number being settled by a council according to the circumstances of the people. The girl's family must also return any buffaloes given as podri. According to Harkness the fines were in his day much heavier; three buffaloes when the man annulled the marriage, and as many as fifty when this was done by, the woman (see P. 538), and the Todas acknowledge that the fine for refusing to fulfill the marriage contract is now lighter than it used to be.

When a girl goes to join her husband she may be given clothing or ornaments by her parents or brothers, and their gifts are known as adrparn or dowry, but I could not learn that there were any definite regulations prescribing what should be given. It seemed also that occasionally buffaloes might be given as adrparn.

THE REGULATION OF MARRIAGE

The Todas have very definite restrictions on the freedom of individuals to marry. One of the most important of these is that which prevents intermarriage between the Tartharol and the Teivaliol. These groups are endogamous divisions of Toda people. Although a Teivali man is strictly prohibited from marrying a Tarthar woman, he may take a woman of this division to live with him at his village, the man being known as the mokhthodvaiol of the woman. This connexion, which will be more fully considered the end of this chapter, may be regarded as a recognised form of marriage, but it differs from the orthodox form in that the children of the union belong to the division of the mother. They do not, however, belong to her clan, but to that of her legal husband. Similarly, the same kind of connexion may be formed between a Tarthar man and a Teivali woman, but in this case the woman is not allowed to live at the village of the mokhthodvaiol, who may either visit her occasionally or go to live at her village.

It has already been mentioned that each of the two divisions of the Toda community is divided into a number of septs or clans, and these are definite exogamous groups. No man or woman may marry a member of his or her own clan, but must marry into another clan. This restriction applies even to the members of clans which are known to have separated from one another in recent times. Thus, among the Tartharol certain members of the Melgarsol separated from the main group, and their descendants have formed a separate group or groups known as the Kidmadol and Karshol, but although the separation took place many years ago there still remains a definite prohibition against a marriage of members of these clans with the Melgars people. The clans of Pedrkars and Kulhern among the Teivaliol are offshoots of the Kuudrol, but here the separation seems to have occurred so long ago that the common origin is not regarded as a bar to marriage.

In the whole of the genealogical record given in the tables at the end of the volume there is not a single case in which marriage has occurred between two members of the same clan. Among many races at or below the stage of culture of the Todas prohibition of marriage within the clan is usually accompanied by prohibition of sexual intercourse, and such intercourse is regarded as incest and often as the greatest - of crimes. . It is doubtful whether there is any such strict prohibition among the Todas. In the qualifying ceremony for the office of palol known as tesherst, it is ordained that the woman who takes part in the ceremony shall be one who has never had intercourse with one of her own clan, and I was told that it was far from easy to find such a woman. The fact, however, that this restriction should exist in connexion with a ceremony suggests that even to the Todas there is something reprehensible in intercourse between man and woman of the same clan (see also P. 530).

There are certain special prohibitions against marriage between members of certain clans. Among the Tartharol the Panol are not allowed to marry the Kanodrsol, a prohibition said to be due to the murder of Parden by Kwoten, and it is said that since that day no marriage has ever taken place between the clans of the two men. In the genealogical record there is no case in which these two clans have intermarried.

I was also told that the people of Melgars and Kwodrdoni might not intermarry, but there are three examples of such marriages in the genealogies. I could not obtain any reason for the restriction, and the information is probably incorrect. The restrictions on marriage between the people of Melgars and those of Kidmad and Karsh have already, been considered.

Among the Teivaliol there are also prohibitions against intermarriage between certain clans. The people of Piedr may not marry those of Kusharf. judging from the genealogical record, the prohibition is not strictly followed, for three such marriages have taken place in recent times. In one of these cases, however, in which a Piedr man married a Kusharf woman, the woman soon became seriously ill, and the marriage was annulled. I could obtain no reason for the prohibition of marriage between these two clans. Marriage was also prohibited between the Piedrol and the Pedrkarsol, this being due to a comparatively recent quarrel between members of the two clans, of which an account is given in Chapter XXVIII.

I have analysed the genealogical records with the view of ascertaining whether certain clans intermarry with any special frequency. Among the Tartharol, I find that the people of Nodrs marry most frequently those of Kars and Taradr. The Karsol, the largest of the Tarthar clans, distribute their marriages widely over the whole Tarthar division. The Panol chiefly marry with Kars and Melgars.

The Taradrol have married most often with Nodrs, Kars and Melgars Keradr, a very small clan, shows no special predilection. The people of Kanodrs have intermarried most often with Kwodrdoni, Pam, Kars and Melgars. The people of Kwodrdoni marry most often people of Kanodrs, Kars and Nidrsi. The Pamol have married chiefly with Kanodrs, Kars and Melgars. Most of the Nidrsi marriages have been with Kars. The Melgarsol have married in fairly equal proportions people of Nodrs, Kars, Taradr and Pam. These facts are interesting in that they show that there is a tendency for the three clans of Nodrs, Kars and Taradr to intermarry. These are not only the most important Tarthar clans, but they occupy the same district of the hills, in the centre and towards the north and north-west, Similarly, the clans of Kanodrs, Kwodrdoni and Pam, situated towards the north-east and east, show a distinct tendency to intermarry. Further, the Melgarsol, who form a special group standing somewhat apart from the rest, distribute their marriages fairly equally, but have often married with Pam, a clan seated near them geographically. The analysis of the genealogies shows that the geographical distribution of the Tartharol on the hills has had a definite influence on the intermarriage of the different clans.

Among the Teivaliol, intermarriage has been greatly influenced by the enormous size of the Kuudrol as compared with the other clans of the division. In order to marry outside their own clan, the people of Kuudr have married nearly all the available members of the other clans of the Teivaliol, leaving very few to intermarry with one another. Thus the genealogies record 161 marriages between Kuudrol and members of the other five Teivali clans, leaving only sixteen marriages between the members of those five clans. Owing to the enormous development of one clan, the Teivali division has almost come to be in the position of a community with a dual marrying organisation in which every member of one group must marry a member of the other group, but there is no reason whatever to think that this is due to any other reason than the excessive development of one clan in numbers.

On studying the marriages in detail, it is found that the Kuudrol have married members of the Piedr clan most frequently, but this is chiefly because the Piedrol stand second to the Kuudrol in point of numbers, although it is also furthered by the restriction in marriage between Piedr and Kusharf. The marriages of the Kuudrol with other clans seem to be determined more by the numbers available than by any predilection for special clans. Both Pedrkars and Kulhem are said to be offshoots of the Kuudrol, but apparently the separation is so remote that the common origin is not regarded as a bar to marriage. it is possible that the necessity of providing spouses for the Kuudrol has tended to break down a restriction which probably once existed.

The Todas have never married people outside their own community, and a strong prejudice against such marriages still exists. This may be illustrated by two recent cases.

A woman, married in the usual way, was divorced by her husband because she became ill. She returned to her own home, where she was visited by a Tamil blacksmith. The latter was very anxious to marry the woman and on one occasion took her away to the plains, but she was followed by her relatives and brought back to her home. Later she married two Toda brothers and was taken by them to their village, but she was followed by the blacksmith, who brought her back to the village of her parents. The Todas seem to have no strong objection to her relations with the stranger so long as she remains among themselves.

In the other case a woman about twelve years ago was visited by a rich Mohammedan who gave money to her husbands, and it was said also that he bribed the chief Toda people, ie., the members of the council. The Mohammedan wished very much to marry the woman and for a sum of money the Todas consented. After the woman had lived for a few days in the bazaar with her new husband, her relatives came and took her away, and I was told that the Mohammedan took the loss so much to heart that he died of grief, but my informants were doubtful whether his grief was due to the loss of his wife or whether it was because he had impoverished himself by the bribes which he had given. Here again the people appear to have had no objection to the relations of the woman with the Mohammedan so long as she remained in the community.

Kinship and Marriage

The members of his own clan are not the only kin whom a man is not allowed to marry. The Todas have a general term, piiliol, for those relatives whose intermarriage is prohibited. The term is applied by a man not only to the women whom he may not marry, but also to the families in general into which he may not marry; thus a man may speak of other men as his piiliol, meaning by this that he may not marry their sisters. This, however, is only a loose way of using the word, and, putting on one side this sense with which the word may be used, the following are the piiliol of a man:--

(i.) The daughters of his father's brothers, whom he would call akka or enda, according to age.

(ii.) The daughters of his mother's sisters, also akka or enda.

(iii.) The sisters of his father and conversely the daughters of his sisters, ie., his mumi and his wankugh.

(iv.) The daughters of the sisters of his father's father, i.e., of the sisters of pian.

The relatives under the first head will be members of the same clan as the man, and the prohibition of marriage between piiliol under this head may be regarded as a restriction on either clanship or kinship.

There seemed to be no doubt, however, that in connection with marriage, a man always thought of these relatives as piiliol included all the people of his own clan. If I am right in this, it means that it is the bond of blood-kinship which a Toda has chiefly in his mind when he considers whether he may or may not marry a given woman.

He has not two kinds of prohibited affinity, one depending on clan relations, and another on relations of blood-kinship, but he has only one kind of prohibited affinity, to which he gives the general term joidiel, including certain kin through the father and certain kin through the mother, and there is no evidence that he considers the bond of kinship in one case as different from the other as regards restriction on marriage. The fact that the Toda includes all those kin whom he may not marry under one general term, and that the kin in question include members both of his own and other clans, goes to show that the Todas recognise the blood-kinship as the restrictive agency rather than the bond produced by membership of the same clan.

The analysis of the genealogical record has shown that these restrictions on marriage are enforced. I have already stated that the genealogies show no single case in which marriage has occurred between members of the same clan. i.e., between piiliol who come under, the first head in the list given above.

I have also failed to find a single case in which marriage has taken place between the children of two own sisters, or of marriage between the children of two women who would call each other "sister" whose names occur in the same genealogical table. Thus I have found no case in which a marriage has taken place between the children of two women so closely related to one another as Punzueleimi and Nasturs, of Table 3, these women being first cousins according to our system of kinship.

It would be a prolonged task to ascertain whether marriage ever takes place among the Tartharol between the children of two clan-sisters in the widest sense, and I do not know whether such marriage may not sometimes occur.

Among the Teivaliol marriages between clan-sisters even in the widest sense must be very rare owing to the fact that nearly all marriages take place between people of Kuudr on the one hand and members of the five other Teivali clans on the other. Since in most cases two women of any one of these five clans marry men of Kuudr, marriage between their children would be restricted under the first prohibition, and similarly the children of two Kuudr women could only intermarry in those cases in which members of the other five clans have married one another. Among the Teivaliol, I do not believe that marriages take place between the children of sisters in the widest sense, and I have little doubt that they are very exceptional among the Tartharol.

There is no case in the genealogies in which the third restriction has been broken, in which a man has married his father's sister or his sister's daughter, his mumi or his mankugh.

There is at least one case in the genealogies in which there has been an infringement of the fourth restriction given on page 50. The marriage of Nargudr (62) with Tolveli (58) is an example of the marriage of a man with the daughter of his grandfather's sister. I believe that this restriction is part of a wider regulation. Using Toda terms of kinship the law would run: a person must not marry the child of his matchuni. The marriage of a man with the 'daughter of his grandfather's sister, such as that of Nargudr with Tolveli, would be an infringement of this law. I have only found one other case in the genealogies in which this law would have been broken, ie., in the marriage of Teitnir (52) and Tersveli (63). Tersveli's father, Teikudr, is the son of Kavani, the sister of Pareivan, Teitnir's father. Teikudr is therefore the matchuni of Teitnir, who has married his daughter.

I was told that though a man might not marry the daughter of his sister, he might marry the children of this woman. I do not know of any such marriage and it is improbable that it would often come about, since it would involve the marriage of a woman with the brother of her grandfather. There is, however, at the present time an example of the marriage of a woman with her father's mother's brother, whom she would therefore call pia, or grandfather. This is the marriage of Kaners and his brother Kudrievan (63) with Edjog (56), the daughter of Toliners, the son of the sister of the two men. I was told, however, that this marriage met with a good deal of disapproval among the Todas, but I could not learn that there was any definite prohibition against it.

The Marriage of Matchuni

While marriage with the daughter of a father's brother and a mother's sister is prohibited, the daughter of a father's sister or a mother's brother is the natural wife of a man. The orthodox marriage is marriage between matchuni, the children of brother and sister. Thus it is obviously not nearness of blood-kinship in itself which acts as a restriction on marriage, but nearness of blood-kinship of a certain kind.

I have analysed the genealogies to ascertain the frequency with which marriages between matchuni occur. The genealogical tables record about 550 marriages, of which 373 are Tarthar and 177 Teivali. Only a small proportion of these are marriages between children of own brother and sister. Among the Tartharol there are 40 and among the Tei-valiol 25 such marriages, making together 65 or 11-8 per cent.

Since, however, the matchuni of a man include a much wider circle of relatives than the children of his mother's own brother and father's own sister, the number of marriages between matchuni is very much larger than this.

Nearly all the Teivali marriages are marriages between matchuni in this wider sense, while among the Tartharol there are also many other marriages of this kind.

One of the reasons why the orthodox marriage custom is not still more commonly followed is the existence of the practice called terersthi, to be considered later in this chapter. According to this practice wives are transferred from one man to another, and in this transference no attention appears to be paid to the kinship tie. The woman, or rather girl, originally married to a man may have been his matchuni, but the woman who finally becomes his wife by the working of the terersthi custom may not be and probably in most cases is not his matchuni. In many cases in the genealogies, the original infant marriage may have been forgotten, and the marriage recorded may be the result of the terersthi custom. If I had a complete record of all infant marriages, I have no doubt that the proportion of marriages between matchuni would have been larger.

In some families marriages between matchuni in the near sense occur much more frequently than in others. Thus of the forty matchuni marriages among the Tartharol, the husband or wife belonged to the Taradrol in fifteen cases, and in one large Taradr family, that of Parkeidi (21), six out of eight children married their matchuni in the near sense. It is perhaps significant in this connexion that the Taradrol have been comparatively little affected by outside influences. They are a clan which might be expected to keep up the orthodox Toda custom.

Another example of a family in which the orthodox marriage custom has been frequently followed is that of Table 52, where there may be found eight cases of the marriage of matchuni in the near sense, and several others where the matchuni relationship is more distant.

In some cases marriages have taken place between the children of matchuni. Thus the marriage of Uvolthli 65) with Sinmundeivi (20) among the Tartharol, and of Pangudr (66) with Nelbur (54) and Kanokh (56) with Sanmidz (63) among the Teivaliol, are all cases in which marriages have taken place between the children of two men who called one another matchuni. There may be other cases, but these examples are perhaps sufficient to show that these marriages may be held to take the place of the orthodox matchuni union.

While marriages between matchuni are the rule and marriages between the children of matchuni certainly not unlawful, we have seen that marriage with the child of a matchuni is prohibited. From our point of view, this means that while marriage with a first cousin is orthodox, marriage with a first cousin once removed is unlawful, while again it seems that marriage with a first cousin twice removed may be lawful. The more distant tie of kinship from our point of view is unlawful, while the nearer is commanded.

Marriage with a matchuni may often involve considerable disproportion of age. In one case at the present time a boy of about two years of age is married to a woman of about twenty. The woman, Nulnir (10), was still unmarried when she reached this age, so she was married to her matchuni -- Kagerikutan (25), the son of her mother's brother. In this case the orthodox marriage was resorted to when the woman had failed to obtain a husband in any other way, although it involved marriage with a baby.

In another case, the marriage of Keitkarg (38) and Potoveli (49), in which the woman is considerably older than her husband, the husband and wife are matchuni.

There is one ceremonial marriage in which the husband always stands in the relation of matchuni to the wife. This is in the performance of the pursutpimi ceremony at the funeral of a girl unmarried at the time of her death. The boy who is chosen to give the bow and arrow and to act as the husband is always, so far as I could discover, the matchuni of the dead girl.

Similarly, if an unmarried boy dies, the girl who is chosen to act as his widow should be his matchuni. In one case of which I have a record, the son of Tutners (58) died and Sotidz (66) was chosen to act as widow. None of the brothers of Puvizveli (65), the mother of the dead boy, had at that time a son, so the duty was undertaken by the daughter of Pangudr, of the same clan as Puvizveli, but belonging to a different family. In this case the matchuni was the daughter of a clanbrother because there was no nearer matchuni available.

Keinba, who acted as husband at the funeral of Sinerani (see P. 394), was the matchuni of the dead girl in two ways, as the son of her mother's brother and as the son of her father's half-sister.

A matchuni may be either the child of a mother's brother or of a father's sister, and I have examined the genealogies to see if a man marries the daughter of his mother's brother or of his father's sister the more frequently, and find that there is no great difference, though the former marriage is somewhat the more frequent. There are among the Tartharol twenty cases in which a man has married the daughter of his mother's brother, two of marriage with the daughter of a stepmother's brother, and one with the daughter of a stepmother's half-brother, making twenty-three cases in all. On the other band, a man married the daughter of his father's sister in fourteen cases, twice he married the daughter of his father's half-sister, and once the stepdaughter of his father's sister, making seventeen cases in all.

Among the Teivaliol marriages with the daughter of a father's sister are the more frequent, there being fifteen of these as compared with ten cases of marriage with the daughter of a mother's brother. There is evidently n*o special preference for either kind of marriage.

POLYANDRY

The Toda have a completely organised and definite system of polyandry. When a woman marries a man, it is understood that she becomes the wife of his brothers at the same time. When a boy married to a girl, not only are his brothers usually regarded as also the husands of the girl, but any brother born later will similarly be regarded as sharing his older brothers' rights.

In the vast majority of polyandrous marriages at the present time, the husbands are own brothers. A glance through the genealogies will show the great frequency of polyandry, and that in nearly every case the husbands are own brothers. In a few cases in which the husbands are not own brothers, they are clan-brothers, ie., they belong to the same clan and are of the same generation. Instances of such marriages are those of Toridz (65) with Kulpakh (52) and Kiladrvan (60), and of Sintharap (68) with Kuriolv (52) and Onadj (57).

There is only one instance recorded in the genealogies in which a woman had at the same time husbands belonging to different clans, viz., the marriage of Kwelvtars (60) with Nidshtevan of Piedr (64) and Tiftners of Kusharf (67), and in this case the men were half-brothers by the same mother, the fathers being of different clans. While I was on the hills, there was a project on foot that three unmarried youths belonging to three different clans should have a wife in common, but the project was frustrated and the marriage did not take place.

It is possible that at one time the polyandry of the Todas was not so strictly 'fraternal' as it is at present, and it is perhaps in favour of this possibility that in the instance of polyandry given by Harkness the husbands were obviously not own brothers. It must be remembered, however, that this case came to the notice of Captain Harkness because the polyandry had led to disputes, and, as we shall see shortly, it is in those cases of polyandry in which the husbands are not own brothers that disputes arise.

The arrangement of family life in the case of a polyandrous marriage differs according as the husbands are, or are not, own brothers. In the former case it seemed that there is never any difficulty, and that disputes never arise. The brothers live together, and my informants seemed to regard it as a ridi- culous idea that there should ever be disputes or jealousies of the kind that might be expected in such a household. When the wife becomes pregnant, the eldest brother performs the ceremony of giving the bow and arrow, but the brothers are all equally regarded as the fathers of the child. If one of the brothers leaves the rest and sets up an establishment of his own, it appeared, however, that he might lose his right to be regarded as the father of the children.

If a man is asked the name of his father, he usually gives the name of one man only, even when he is the offspring of a polyandrous marriage. I endeavoured to ascertain why the name of one father only should so often be given, and it seemed to me that there is no one reason for the preference. Often one of the fathers is more prominent and influential than the others, and it is natural in such cases that the son should speak of himself as the son of the more important member of the community. Again, if only one of the fathers of a man is alive, the man will always speak of the living person as his father; thus Siriar (2o) always spoke of Ircheidi as his father, and even after Ircheidi is dead, it seems probable that he will so have fallen into the custom of speaking of the latter as his father that he will continue to do so, and it will only be when his attention is especially directed to the point that he will say that Madbeithi was also his father.

In most of the genealogies, the descent is traced from some one man, but there can be no doubt whatever that this man was usually only one of several brothers, and the probable reason why one name only is remembered is that this name was that of an important member of the community, or of the last surviving of the brother-husbands.

When the husbands are not own brothers, the arrangements become more complicated. When the husbands live together as if they were own brothers there is rarely any difficulty. If, on the other hand, the husbands live at different villages, the usual rule is that the wife shall live with each husband in turn, usually for a month at a time, but there is very considerable elasticity in the arrangement.

It is in respect of the 'fatherhood’ of the children in these cases of non-fraternal polyandry that we meet with the most interesting feature of Toda social regulations. When the wife of two or more husbands (not own brothers) becomes pregnant, it is arranged that one of the husbands shall perform the ceremony of giving the bow and arrow. The husband who carries out this ceremony is the father of the child for all social purposes; the child belongs to the clan of this husband if the clans of the husbands differ and to the family of this husband if the families only differ. When the wife again becomes pregnant, another husband may perform the pursutpimi ceremony, and if so, this husband becomes the father of the child; but more commonly the purutpimi ceremony is not performed at all during the second pregnancy, and in this case the second child belongs to the first husband, i.e., to the husband who has already given the bow and arrow. Usually it is arranged that the first two or three children shall belong to the first husband, and that at a succeeding pregnancy (third or fourth), another husband shall give the bow and arrow, and, in consequence, become the father not only of that child, but of all succeeding children till some one else gives the bow and arrow.

The fatherhood of a child depends entirely on the pursutpini ceremony, so much so that a dead man is regarded as the father of a child if no other man has performed the essen- tial ceremony.

In the only case in the genealogies in which the husbands of a woman were of different clans, it happened there were only two children, and that one father gave the bow and arrow for the first child and the other for the second. If the husbands separate, each husband takes with him those children who are his by virtue of the pursutpimi ceremony.

There is no doubt whatever as to the close association of the polyandry of the Todas with female infanticide. As we have seen, the Todas now profess to have completely given up the practice of killing their female children, but it is highly probable that the practice is still in vogue to some extent. It has certainly, however, diminished in frequency, and the consequent increase in the proportion of women is leading to some modification in the associated polyandry.

It has been stated by most of those who have written about the Todas that the custom of polyandry is dying out, but a glance at the genealogies will show that the institution is in full working order even in the case of the infant marriages which are being contracted at the present time. There is, however, some reason to believe that it is now less frequent for all the brothers of a family to have one wife only in common. A study of the genealogies shows that often each brother has his own wife, or that several brothers have more than one wife between them. It seemed to me, however, almost certain that in these cases the brothers have the wives in common. In compiling the genealogies, one informant would give me the names of two or more brothers each with one wife, while another would give me the name of one brother with two or three wives, and would say that the other brothers had the same wives. When I pointed out the discrepancy and asked which was the true account, they usually said it made no difference and were almost contemptuous because I seemed to think that there was any disagreement between the two versions. I think it probable that it has become less frequent for several brothers to have only one wife in common, but I am very doubtful whether this indicates any real decrease in the prevalence of polyandry.

It seems to me that the correct way of describing the present condition of Toda society is to say that polyandry is as prevalent as ever, but that, owing to the greater number of women, it is becoming associated with polygyny. When there are two brothers it does not seem that each takes a wife for himself, but rather that they take two wives in common.

It is probable that this will lead in time to a state of society in which each brother will come to regard one wife as his own; and in a few cases it seemed to me that there was already a tendency in this direction. If this forecast should be fulfilled, the custom of monogamy among the Todas will have been developed out-of polyandry through a stage of combined polyandry and polygyny.

One case happened during my visit which seemed to indicate that though several brothers might be regarded as husbands of a woman, the part of husband for ceremonial purposes might be taken only by one or two of them. In this case I was told that four brothers had one wife, but when the wife died only two of the brothers acted as widowers and performed the ceremonies associated with that condition. When I asked for an explanation of this, I was then told that the other two brothers were not husbands, but I strongly suspected that this was a mere device to enable two of the brothers to avoid the disabilities attendant on the condition of widowerhood. I have very little doubt that while the woman was alive, all the four brothers were her husbands, but after her death it became convenient to assume that only two had been husbands, leaving the others free from the restrictions of widowerhood.

Many writers have believed that the widely spread custom of the levirate is a relic of polyandry. If it were true that the custom of polyandry is dying out among the Todas, this people might have provided material for the study of the relations of polyandry and the levirate. It will be obvious, however, from the account already given, that polyandry is still strongly established among the Todas. Still, there are a few cases in the genealogies which seem to show that when two brothers had different wives, and one brother died, the widow might be taken by the surviving brother. Thus, in Table 34, two brothers, Matovan and Kemners, had one wife, Sargveli, while Atcharap had his own wife, Puners. When Matovan died, Sargveli was regarded as the wife of both Atcharap and Kemners. Again, after the deaths of Mulpolivan and Peigvan (3), the widow of Nersveli was married by Pero, the clan-brother (first cousin) of the husband.

In other cases, the widow of one brother has not become the wife of her husband's brothers, but has married elsewhere; and though the evidence is necessarily very unsatisfactory, it seems on the whole probable that the Todas show no special relation between polyandry and the levirate custom.

If the widow marries a man who is not one of the brothers of her dead husband, the new husband has to pay a certain number of buffaloes. He does not, however, give these buffaloes to the brothers of the dead man, but to his children; thus, when Karnisi of Pdm (37) died, his widow, Nersaveli, married Mutthuvan (34) of Kan6drs, who paid fourteen buffaloes to Pungievan, the son of Karnisi. This payment of buffaloes is known as terkudrichti, "compensation he gives," and it is the custom for the number of buffaloes in this case to be twice the number given by the dead man for- his wife; in this case Karnisi had taken Nersaveli from another man for seven buffaloes.

In relation to the Levirate, the important point here is that the buffaloes are paid to the sons of the dead husbands, not to his brothers.

I do not think that the Todas provide any definite evidence towards the solution of the vexed question of the relation between polyandry and infanticide. It is possible that at their first arrival in the Nilgiri Hills, the Todas had few sources of food, and had a severe struggle for existence; that they therefore adopted the practice of female infanticide, and that polyandry followed as a consequence. At the present and during recent times there has certainly been no economical motive for infanticide, and I am very doubtful whether it has ever existed. I think it far more probable that the Todas brought the practice of polyandry with them when they came to the Nilgiris ; but if this view should be adopted, there is still no evidence to show whether they also brought infanticide with them, or whether this custom developed owing to the fact that polyandry diminished the need for female children.

Polygyny

In the last section we have seen that there is a tendency for the polyandry of the Todas to become combined with polygyny. Two brothers, who in former times would have had one wife between them, may now take two wives, but as a general rule the two men have the two wives in common. In addition, polygyny of the more ordinary kind exists among the Todas, and is probably now increasing in frequency, as one of the results of diminished female infanticide.

One example of polygyny is the marriage of Kuriolv (56) with two wives, one of about the same age as himself, the other a young wife whom he shares with Onadj (57). In another case Odrkurs (1), has two wives, the second wife being a young girl recently married in the hope of obtaining a son (see P. 550).

There is one example of polygyny in the genealogies in which a young boy, Mokudr of Nidrsi (42), has two wives, both young girls. He has been doubly married in order that he may get rid of one of his wives by the terersthi custom and so become rich. He has been married to two wives in order that he may sell one.

When a man or a group of men have more than one wife, the two wives usually live together at the same village, but sometimes they live at different villages, the husband or husbands moving about from one village to the other.

EXCHANGE OF BROTHER AND SISTER

Although I was not told that it was the custom for a brother and sister of one clan to marry a sister and brother of another, examination of the genealogies makes it clear that this frequently happens. A good example which may be cited is the marriage of Kuriolv (52) with Punaveli (65) while Sink6rs, the sister of Kuriolv, married the three brothers of Punaveli. Two other similar instances may be found in Table 52, and they are of general occurrence throughout the genealogies.

In some communities this custom of exchange is definitely connected with the bride-price, which may be so large as almost to compel a man to give his sister in exchange for the wife he takes from another clan. In the case of the Todas the bride-price is so inconsiderable that it is unlikely that it would form a motive for exchange, and I think it improbable that in such marriages as those cited above, the idea of exchange is even definitely formulated, but that the combination of marriages, comes about for such obvious reasons as may occur in any community. The marriage of matchuni, if widely practised, would obviously lead to an appearance of exchange, and it may be that among the Todas this is the chief cause of its occurrence. Similarly, instances will be found in the genealogies of two brothers (or two groups of brothers) marrying sisters. An example may be given from Table 53, where Orzevan marries one woman and his two brothers marry her sister. Another instance may be found in Table 58.

In several cases in which a man or group of men have had two wives the wives have been related. Thus, Kutthurs (12) and his brothers first married Tedjveli (16). After her death, Kutthurs, the only surviving brother, married Sabnir (34), the daughter of Arsner, Tedjveli's sister. Again, Paners (23) and his brothers first married Pergveli, and when she died the), married her brother's daughter. Pungusivan (53) married his matchuni, Sinodz (68), and when she was taken from him by the 4erersthi custom, he married Sintharap, her sister.

There is often very great disproportion of age in Toda marriages. I have already given two cases in which the woman is the older, in each of which the disproportion of age is due to the custom of marrying a matchuni. More commonly the man is much the older ' and there are at the present time many cases in which elderly men are married to young girls. This is partly due to the practice of infant marriage. Unless a widower can take advantage of the terersthi custom, which is always expensive, he may have to marry a child and wait till she has reached a marriage. Thus, Kodrner, my guide, lost his wife some years ago, and then married a girl whose present (1902) age is only thirteen, Kodrner being forty-two. The girl is still living with her parents, and will probably not go to her husbaid for another three or four years.

THE CUSTOM OF 'TERERSTHI'

The marriage tie among the Todas at the present time has become very loose. Wives are constantly transferred from one husband, or group of husbands, to another, the new husband or husbands paying a certain number of buffaloes to the old. The amount of the compensation or ter is settled by a council, and from this the transaction has received its name of terersthi, or "compensation he tells (decides)."

There is much reason to believe that this custom has altered its character in recent times. I was told that formerly the custom only applied to cases in which a man had lost his wife by death. If he wished to marry a woman who was already the wife of another or others, he went to the father of the woman and asked for his consent. The father would consult with two other elders, and if they were in favour of the proposed transaction the three elders would go to the woman, and if they obtained her consent they then, went to her husband for his. If husband or wife were unwilling to be parted nothing was done, but if both consented, the new the old husband, the father of the woman, and the two elders met and decided on the number of buffaloes to be given as ter or compensation. This meeting was called terersthi. The ter had to be paid within a month, and all the buffaloes given had to be females. The man who was giving up the woman went to the village of the new husband and received his buffaloes, of which he was allowed to choose a certain number. If he had been awarded more than four buffaloes, he might choose three, if four or less, he might only choose two. Among the Tartharol, a man would usually choose wursulir, and among the Teivaliol, tashir.

At the present time the number of buffaloes given as ter varies very greatly; the most frequent number is three, but often more are given, and in one case,, about ten years ago, a man had to give twenty-five. The number seems to depend largely on the size of the herd possessed by the man taking a new wife. The more buffaloes he has, the more he has to pay.

When the buffaloes are given, the new husband has to give a feast, after which the old husband drives away his buffaloes. In a recent case Teigudr of Nodrs (4) had taken Uwer frorn Nertolvan and Palpa of Pan (16) for nine buffaloes. These two men went to the village of Tedshteiri, where Teigudr was living, and were feasted, the food being cooked on nine ovens, corresponding to the number of the buffaloes. This cor- respondence between the number of the ovens and of the buffaloes given as ter, suggests that there may have been some definite ceremonial in connexion with this feast of which I failed to obtain an account.

The custom of terersthi has some reason on its side. Wherever infant marriage exists in a small community, it must often happen that a widower finds all the women of his community married, and without some machinery by which he is allowed to take the wife of another, he must remain unmarried or be content with marriage to a mere child Even at the present time, we have seen that an adult man who has lost his wife may marry a girl only a few years of age.

At the present time the custom of terersthi has a far wider range. It is obvious that when a widower takes the wife of another he is simply transferring his difficulty, and the man whose wife he has taken will have to seek a new partner. It often happens that a man takes the wife of a boy married, perhaps, to a girl of about the same age as himself, and when this boy reaches manhood he will have to seek a wife and will naturally try to obtain the wife of another rather than be content with a child perhaps only three or four years of age. it would be impossible that such a custom as that of terersthi should remain limited in scope, but there is no doubt that at the present day it has become the custom for any man who takes a fancy for the wife of another to endeavour to obtain her for himself, and I was told that he would give large bribes to the elders of the Todas to attain his object. It seems quite clear that, at the present time, it is not considered necessary to obtain the consent either of the wife or of the husband, and in some cases the wife has been taken from her husband by force. In some recent cases the aggrieved parties in such disputes have appealed to the Government, and during my visit a petition was being drawn up for presentation to the Governor of Madras' asking that the abuses of the terersthi custom should be remedied.

DIVORCE

Divorce exists among the Todas quite apart from the transference of wives just considered. I was told that a man divorces his wife for two reasons, and for two only, the first reason being that the wife is a fool and the second that she will not work. Barrenness is not generally regarded as a reason for divorce, though I was told of one case in which a man had sent away his wife on this account. It seemed more usual in such a case to take a second wife. In some cases the illness of the husband has been regarded as a ground for divorce. Intercourse between a wife and another man is not regarded as a reason for divorce but rather as a perfectly natural occurrence.

When a man divorces his wife, the woman's people usually complain to the naim or council, but if it is decided that the man shall take his wife back, there appears to be no way of compelling him to do so. In any case the husband pays a fine (kwadr) of one buffalo to the wife's people, just as he would have done if he had refused to take her when she reached the marriageable age, but he receives back any buffaloes he may have given as podri. Even if the council decides that the man ought to take his wife back and he refuses, a fine greater than one buffalo cannot be inflicted. If the divorced woman re-marries, the previous husband does not receive anything, and any buffaloes given become the property of the woman's family.

THE MOKHTHODITI INSTITUTION

In addition to the regular marriage, there is another recognised mode of union between men and women, which is called mokhthoditi. The man who becomes the consort of a woman in this way is called her mokhthodvaiol -viz. "man who keeps -mokh," and the woman is called sedvaitazmokh-viz., " woman who joins." The mokhtlioditi union differs from the regular marriage in one important respect. It may be, and usually is, formed between Tarthar men and Teivali women, or between Teivali men and Tarthar women. The great majority of instances of which I heard were of this kind. One woman might have more than one mokhthodvaiol, the largest number of which I heard being three. Similarly, a man might have more than one sedvaitamnokh, but as the custom entailed considerable expenses on the man, this was not common, and I did not hear of any instance in which a man had more than two.

The mokhthodvaiol has no rights over any, children who might be supposed to be his; they are regarded as the children of the regular marriage, This would be the case even if the husband were dead or separated from his wife. If a Teivali man took a Tarthar widow as sedvaitazmokh, and a child were born, the child would belong to the Tartharol, and would be regarded as the son of the dead husband of the woman, and would belong to his clan. The child might live with the Mokhthadvaiol, and be spoken of ordinarily as the child of this man, but yet for all social and legal purposes, the child would be a member of its mother's husband's clan. The dead husband is regarded as the father because it was he who last performed the pursatpimi ceremony.

There are two forms of the mokhthadili union. the woman lives with the man just as if she were his real wife, almost the only difference being thatany children would be legally the children of the legal husband of the woman or sf some man of her division called upon to perform the pursutpimi ceremony. In the other and more usual form the man visits the woman at the house of her husband.

Owing to the restriction on the visits of Teivali women to Tarthar villages, there is a difference in the nature of the mokhthoditi union in the two divisions. A Teivali mokhthodvaiol may take his wife to live with him at one of the Teivali villages, but in those cases in which Tarthar men live permanently with Teivali women, the moklithodvaiol must live at the woman's village. There are two examples of this practice at the present time in which Tarthar men live altogether at Teivali villages.

When a man wishes to have a given woman as his sedvaitazmokh he goes to the husband or husbands of the woman and asks for his or their consent. As a sample of the kind of negotiations which ensue, I will give a definite instance.

A Tarthar man wished to become mokhthodvaivl to the wife of two Teivali brothers. He went to them and asked for their consent, which they gave, but said they should like to have the agreement confirmed by a third party (nedrvol), and they settled on a nedrvol to whom all went. The nedrvol asked each if he consented to the arrangement, and it was decided that the Tarthar man should give a putkuli worth three rupees annually to the woman's husbands, and the former became mokhthodvaiol to the woman on that day.

A few days later the two husbands and the mokhthodvaiol went to the woman's father and brothers (called collectively paiol), and the mokhthodvaiol promised that he would give the woman either a keivali (necklace) or a sin (gold earrings), each worth about thirty rupees. [A poorer man might only give a pult1ii (bracelet), worth about twelve rupees]. He also promised that he would give a three-year-old buffalo to the son of the woman, this being called mokh ir kwadrtil i.e., " son buffalo he gives." After making these promises, the mokhthodvaiol performed the salutation of kahnelpudithti to all the paiol, i.e., he bowed down before each, and placed his head beneath their feet. As we have seen earlier, not only are the relatives of the sedvaitazmokh called paiol, the term in use for the relatives of a real wife, but the father of the woman is called mun and her mother mumi, names which are also terms of bloodrelationship. When a man or woman dies, the moklitliodvaiol of the woman and the sedvaitazmokh of the man have definitely assigned duties at the funeral ceremonies. Each wears a ring on the ring finger of the left hand and has to put various things with the left hand into the pocket of the putkuli of the dead person.'

The mokhthoditi institution was first described by Ward in 1821 the man being called by Ward the coombhal (the kumbliol, cloak or blanket man). This is the Badaga name, and it has usually been adopted by those who have since referred to the institution. The custom is said to have originated with the god Kulinkars, who was the mokhthodvaiol of the goddess Nbtirzi, but I could obtain no details of the way in which the custom is supposed to have arisen. The ceremonial connected with the process of becoming a moklahodvaiol is very much like that of the real marriage.

A garment is given or promised and the salutation of kalmelpudithti is paid to the woman's relatives. The chief difference is that the gifts are more numerous and expensive for the mokhthodvaiol than for the husband. Further, in some cases the sedvaitazmokh of a Teivali man may live with him exactly in the same way as a wife.

Except for the prohibition against Teivali women living at Tarthar villages, and the important difference in the mode of descent of the children there seems to be little essential difference in some cases between the mokhthoditi union and marriage. In describing the institution, one of my informants laid great stress on the disability of a man of one division to perform the pursiitpimi ceremony for a woman of the other division and treated this as the essential point of difference. He seemed to regard this ceremonial disability as primary and the other differences as the secondary results, but I do not know how far this is the general Toda view.

SEXUAL MORALITY

From the foregoing account it appears that a woman may have one or more recognised lovers as well as several husbands. From the account given of the dairy ritual, it appears that she may also have sexual relations with dairymen of various grades that, for instance, the wursol, on the nights when he sleeps in the hut, may be the lover of any Tarthar girl.

Further, there seems to be no doubt that there is little restriction of any kind on sexual intercourse. I was assured by several Todas not only that adultery was no motive for divorce, but that it was in no way regarded as wrong. It seemed clear that there is no word for adultery in the Toda language. My interpreter, Samuel, had translated the Commandments shortly before my visit, and only discovered while working with me that the expression he had used in translating the seventh Commandment really bore a very different meaning.

When a word for a concept is absent in any language it by no means follows that the concept has not been developed, but in this case I have little doubt that there is no definite idea in the mind of the Toda corresponding to that denoted by our word 'adultery.' Instead of adultery being regarded as immoral, I rather suspected,' though I could not satisfy myself on the point, that, according to the Toda idea, immorality attaches rather to the man who grudges his wife to another. One group of those who experience difficulty in getting to the next world after death are the kashtvainol, or grudging people, and I believe this term includes those who would in a more civilised community be plaintiffs in the divorce court.

In nearly every known community, whether savage, barbarous or civilised, there is found to exist a deeply rooted antipathy to sexual intercourse between brother and sister. In savage communities where kinship is of the classificatory kind, this antipathy extends not only to the children of one mother, but to all those who are regarded as brothers and sisters because they are members of the same clan or other social unit. In some communities, such as those of Torres Straits, this antipathy may extend to relatives as remote as those we call second and third cousins, so long as descent through the male line from a common ancestor and membership of the same clan lead people to regard one another as brother and sister.

It is very doubtful whether this widespread, almost universal abhorrence is shared by the Todas. I was told that members of the same clan might have intercourse with one another, and in a preliminary ceremony for an office a special part was taken by a woman who possessed the qualification that she had never had intercourse with a man of her own clan, and it was said it was far from easy to find such a woman. When I collected this information, it seemed clear that this meant that a woman who, before marriage had belonged to a given clan, had never had intercourse with a man of that clan. But since a woman joins the clan of her husband, and since, marriage taking place at an early age, the woman belongs to her husband's clan from this early age, it has since occurred to me that an alternative explanation of the restriction is possible; though it does not seem to me to be likely. It is possible that what is meant is that the woman should never have had intercourse with any of her husband's clan except those who are properly her husbands. If this explanation were the correct one, the prohibition would seem to be directed against practices resembling communal marriage, and would be interesting evidence in favour of the existence of this type of marriage, since there are no prohibitions against what does not exist nor has ever existed. As I have said, however, I think it very unlikely that the prohibition is to be interpreted in this way, but I regret very greatly that it did not occur to me to inquire carefully into this point on the spot.

So far as I could tell, the laxity in sexual matters is equally great before and after marriage. If a girl who has been married in infancy, but has not yet joined her husband, should become pregnant, the husband would be called upon to give the bow and arrow at the pursiapimi ceremony and would be the father of the child, even if he were still a young boy, or if it were known that he was not the father of the child. I only heard of one case in recent times in which an unmarried girl had become pregnant. In this case a man who was a matchuni of the woman was called in to give the bow and arrow, but he did not regard himself as married to the woman and did not live with her. That some stigma was attached to the occurrence may possibly be shown by the fact that this woman remained unmarried for some years, and then only married a man who was certainly below the general standard of the Todas in intelligence. The child, a daughter, of the woman died soon after birth, so that I had no chance of ascertaining whether the irregularity of her birth would have had any influence on her position in Toda society. If, however, a child is born without the PursatPimi ceremony having been performed, it is called Padmokh and an indelible disgrace attaches to it throughout life.

From any point of view, and certainly from the point of view of the savage, the sexual morality of the Todas among themselves is very low. It is an interesting subject of speculation how far this laxity is the result of the practice of polyandry, for since low sexual morality brings in its train various factors which tend to sterility, we may have here, as Mr. Punnett has suggested elsewhere, a reason why polyandry is so rare a form of marriage. The practice of polyandry must almost inevitably weaken the sentiment of possession on the part of the man which does so much to maintain the more ordinary forms of marriage.

The low sexual morality of the Todas is not, however limited in its scope to the relations within the Toda community. Conflicting views are held by those who know the Nilgiri Hills as to the relations of the Todas with the other inhabitants, and especially with the train of natives which the European immigration to the hills has brought in its wake. The general opinion on the hills is that in this respect the morality of the Todas is as low as it well could be, but it is a question whether this opinion is not too much based on the behaviour of the inhabitants of one or two villages near the European settlements, and I think it is probable that the larger part of the Todas remain more uncontaminated than is generally supposed.

That the Todas are perhaps not so depraved as they are painted is suggested by two considerations. There is little evidence of the existence of many half-breeds. I examined in one way or another over 500 Todas and must have seen nearly the whole of the 800 people who form the Toda population. I saw few who suggested Tamil or Badaga intermixture and only one boy whose appearance suggested European parentage. A more careful examination than I gave might, however, have revealed other suspicious cases, and perhaps in a race which practices infanticide the absence or paucity of half-breeds may not carry much weight.

The other consideration is of a different kind and tends to show not only that the Todas are not so depraved as they are painted, but that they are not so black as they paint themselves. By means of the genealogical record I was able to work out the relationship to one another of forty-three individual suffering from colour-blindness. Since this condition run mainly in the female line, it does not afford very cogent evidence of paternity; but a full examination of my records seems to show that colour-blind men, or rather males of colour-blind families, had colour-blind descendants more often than perhaps might have been expected if the Toda are in practice quite as promiscuous as their social regulation allow them to be. The record of the affinity of the colour blind suggests that in spite of the theoretical promiscuity, the husbands are, in practice, very often the fathers of their children.

A few histories of individuals may be given as examples of the various marriage customs which have been described in this chapter. One of the most married of Toda women is Puvizveli of Kusharf (65). She was married in infancy to Singudr (55), of the same clan as Sinkors, the mother of Puvizveli, and the two were probably the matchuni of one another, though only in a distant way. Puvizveli was taken from Singudr by Madsu and Koboners (58), who gave for her three buffaloes. From them she was transferred to Kangudr of Piedr (62), it being arranged that he should pay seven buffaloes. Soon after joining Kangudr, Puvizveli became ill, and since there is a prohibition of marriages between the clans of Piedr and Kusharf, it was agreed that he pair should separate, and the woman was taken by Utners and Etamudri (58). The eleven buffaloes had never been paid by Kangudr, so Tutners and his brother gave their buffaloes directly to Madsu and Koboners, but only four instead of eleven. All these transactions took place while Puvizveli was still young, but by her new husbands she had a son who died soon after birth. During her second pregnancy, he was taken by Perpakh and Tebkudr (68), who gave six buffaloes. The transference took place before the pursidpimi ceremony had been performed. Perpakh gave the bow and arrow, and the daughter since born is regarded as the child Perpakh and Tebkudr. Puvizveli has also a Tarthar mokhthodvaiol.

Edjog of Kuudr (56) was married in infancy to Nargudr (62), the son of her mother's brothers, and therefore hermatchuni in the nearest sense. From him Kiudners (70 and his two brothers took her for five buffaloes. Kiudners died before the buffaloes had been paid, and Edjog was taken by Mavodriners (65), who arranged to pay the five buffaloes to Nargudr. He did not do so, but after having a son by Edjog, he sent her back to the father, paying a kwadr of one buffalo. So far, Nargudr had not received his five buffaloes, but he now obtained them from Kaners and Kudrievan (63), who took the woman although she was the granddaughter of their sister Narskuti. The marriage met with disapproval among the Todas on this account, though there does not appear to be any definite regulation against such a marriage; and at the time of my visit Edjog, a young woman of about twenty-seven, was still the wife of the two old men, aged about seventy and sixty-seven respectively.

Kuriolv of Kuudr (52) first married Punaveli (65), by whom he had two children. He then took to live with him Pilimurg (7), a Tarthar woman, giving to Pepners (44), the husband of the woman, fifteen buffaloes. Though Pilimurg is only legally his sedvaitazmokh, Kuriolv treats her as a wife She lives at one of the Kuudr villages, while Punaveli lives at another. Pilimurg has had one son, Meilitars, since she has been living with Kuriolv, and Kuriolv always speaks of the boy as his son, though legally he is the son of Pepners, and his name will be found in the genealogies among the children of this man.

Recently Kuriolv has also married Sintharap (68), sharing her with Onadj (57), of the same clan as Kuriolv, but belonging to a different family. Sintharap has had three children, for the first of whom Kuriolv performed thepursi0pimi ceremony, and since no one has performed this ceremony for the succeeding children, they are also regarded as the children of Kuriolv. One of these children was Sinerani, whose funeral ceremonies have been described.

Kuriolv's son Kulpakh (52), married Toridz (65), sharing her with Kiladrvan (6o), of the same clan as Kulpakh, but of a different family. At the first pregnancy Kulpakh gave the bow and arrow, and was regarded as the father of that child and of two- succeeding children who were born while Kulpakh was alive. After the birth of the third child Kulpakh died, and Toriclz has since continued to live with Kiladrvan and has had two more children. Kuriolv, the father of the dead man, succeeded in preventing Kiladrvan from performing the pursutpimi ceremony before the birth of either of these children, and consequently they are regarded as the children of the dead Kulpakh and belong to Kuriolv's division (polm) of the clan and not to that of Kiladrvan. Here, by virtue of the Aursfilpimi ceremony, a dead man is the legal father of two children who are known to be really the sons of his fellow-husband.

In the preceding cases the people belong to the Teivaliol. Among the Tartharol there are similar histories.

Pupidz of Kwodrdoni (35) was married in infancy to two brothers, Kalgeners and Kinagudr, belonging to the same clan as the mother of Pupidz, so that she would probably have called them matchuni, though they were not nearly related. From these boys Pupidz was taken by Patser (26), who gave for her three buffaloes. From Patser she was taken by Siriar (20) for five buffaloes. Some time later Pepob (44) wished to marry Pupidz, but both she and Siriar were unwilling to be separated. Pepob, however, persuaded the council to arrange that he should have the woman for three buffaloes, and soon after five or six men carried off the woman by force, entering Siriar's hut at night.

Two of the men held Siriar while the others carried off his wife, who became pregnant by Pepob, but Siriar, who had been trying to get back his wife, succeeded when she was about at the sixth month. The hand-burning ceremony had already been performed, but Siriar gave the bow and arrow, and is therefore the legal father of the boy born afterwards, although Pepob is known to have been the real father. Siriar had to give Pepob eleven buffialoes, though he had only received three, and had given five to the previous husband.

Nanbarvan of Kars (7) first married Pothenir (47), by whom he had one son. Nanbarvan went to England with a party of Todas, and Pothenir then married Kutadri, Nanbarvan's first cousin. On his return from England, Nanbarvan married Sifidod (38), by whom he had and since that time he has had no wife, though he claims that fraveli, his brother's wife, is also his. There seems to be no doubt, however, that he does not live with his brother in the same way as in most cases of polyandry, and is a wanderer with no regular home of his own, but I could not discover the cause of this.

A dispute about a marriage was in progress while I was on the Hills, which I did not understand completely, but it appeared that Oselig (24), who had been first married to Teigudr (4), was then taken by Punog (14). Punog was said to have treated his wife badly, and to have failed to perform his duties when therewas a funeral in the family of Nertiners, the brother of Oselig. He had not given the proper pddri, nor had he taken part in the cloth-giving ceremony, so Oselig ran away from him and took refuge with her brother. After a month Punog demanded back his wife and also twelve buffaloes which he had left with Nertiners for grazing purposes. Nertiners refused to send back his wife, and returned only eight of the buffaloes. He also proceeded to arrange that Oselig should marry Udrchovan (36), and Punog accused Nertiners of having got up the whole quarrel in order that Oselig should make this marriage. The matter was referred to the council, and it was decided that Oselig should become the wife of Udrchovan, but I did not hear for how many buffaloes, nor how the other disputes about buffaloes and pddri were settled.

At this time Udrchovan had another wife, Pandut (45). She had been the wife of Udrchovan and his brother Popners from infancy, and after having three children, who died young, she had been sent away and Udrchovan married Kavener (3), while his brother married Silkot (10). Later Kavener was taken from Udrchovan by Kudrvas (i i), and Udrchovan remarried Pandut, who in the meantime had had two other husbands. To the foregoing accounts, which I give as exceptional and not as typical examples of the uncertainty of Toda married life, I add one taken from the book published by Captain Harkness in 1832, p 121. The notes are added by myself.