teristic of the san-andro of the day of burial. If, for instance, it was on Wednesday, the special number of which is 6, they had to stop six times with the bier on the way to the grave, throw down a stone at each stoppingplace, and carry the corpse six times round the grave before they buried ti. If the burial was to take place on a Saturday, the same thing would be done only four times; if on Friday, seven times, and so on, according to the characteristic number of the san-andro of the day.

Why the different planets got these numbers respectively, I am unable to tell. As the Sun has No. 1, it occurred to me at first that the various heavenly bodies were probably numbered according to their supposed distance from the earth; but this cannot be the case. They may have supposed the brightest (the Sun) to have been nearest, and the next brightest (the Moon) to come next; but then we should certainly have expected Venus to follow, and not Mars and Mercury, which are not nearly so bright. It is strange too, that Mars and Mercury should be considered lucky, and the bright Moon, Venus, and Jupiter unlucky. From the little we know of the astrology of the ancients, we gather that the three last ones were considered lucky, Saturn and Mars unlucky, and Mercury neutral.

2.-Ny San-andron' ny Vèlona, na Sa-mivèrina (The San-andro of the Living, or the San-andro which was counted 'backwards').

The description of it given by my native helper is not very clear, but so far as I can make out, this san-andro had reference only to sacrifices (sorona). When a sorona was brought, prayers were to be offered up too, and in these prayers and invocations the priest used to expatiate on the corresponding san-andro; but in so doing he did not refer to the sanandro of the day of the offering, but always to that of the day before yesterday; in other words, he always counted two days backwards to find the san-andro he wanted. If, for instance, Sunday was the day of offering, the san-andro of the preceding Friday was the one he referred to, and so on. It does not fall within the scope of this article to give the prayers which the priests used on such occasions; but this one is so peculiar that I must briefly refer to it. The priest on this occasion used to call on God as 'Andriamanitra fito miànaka' ('God as a family of seven') and as 'válo mivady' ('eight pairs,' i.e. eight husbands and eight wives ?*). He also calls on God as Ratomárajìba, Ratomáraféfy, Rabodìsy and Rakénonkénona. All this is peculiar; the last word is Malagasy and seems to mean 'the loquacious one.' The three preceding ones seem to be foreign words, at least partly so; tomara, bodisy, and jibo seem to be Arabic, but I cannot identify them at present. The 'seven' probably had a reference to to the seven planets as God's manifestation.

Offerings could only be brought on the three 'good days,' Sunday, Tuesday and Wednesday; but sikidy could be performed on any day. The reason probably was, that sikidy belonged to the necessities of life, which could never be allowed to be stopped.

3.-The Character of the Seven Days of the Week in relation to Evils and the Foretelling of Evils.

Both expressions are ambiguous. Fito mianaka may mean parents with five children, or a father, or mother, with six children. Valo mivady most naturally means eight pairs, but it might also mean only four pairs (eight married persons), or even a man with seven wives!

My native professor gives the following rules, which I reproduce on his authority:

1. Alahady was the proper day for everything white: white-haired people (fótsy volo), white stones, etc.

2. Alatsinainy was the day for everything green and blackish: grass, forests, greenish birds, people with blackish skin, etc.

3.

4.

Talata: the day of people who have many scars (cicatrices), and are also pock-marked from the small-pox (sèki-nèndra).

Alarobia: the day of women and everything female.

5. Alakamisy: the day of slaves.

6. Zoma: the day of nobles and everything which is red (red or scarlet clothes, etc., characteristic of the higher nobility).

7. Asabotsy: the day of the young people and everything young.

If a man suffering from or apprehensive of some evil came to a mpisikidy or other foreteller of future events, he would be sure to be asked some question with special reference to the character of the day on which he came. If he, for instance, came on a Sunday (Alahady), it would be intimated that his complaint had been caused by some obnoxious white stone; or by drinking milk (which of course is white), in which there were some ghosts; or that he had been bewitched by some old whitehaired woman; or, at any rate, that he was in danger of some such mishap, and had better look out carefully. If on Thursday (Alakamisy), his griefs were almost sure to be attributed to some slave, or he was warned to beware of his slaves, lest they should murder or bewitch him.

And so on for the other days, according to the nature of the day. 4.-Foretelling of the Tasik andro, (i.e. the day on which one may be in special danger of getting ill through the influence of the vintana.

This was a peculiar compound of vintana and sikidy, subjected to the following rules:

To find the day, you begin from Tuesday (Talata), and work the sikidy on the following principles :

1. Trano and Làlana form (point to) Talata (Tuesday).

2. Làlana and Mpanontàny form Alarobia (Wednesday).

3. Mpanontany and Asorotany form Alakamisy (Thursday).

4. Asorotany and Andriamanitra form Zomà (Friday). 5. Andriamanitra and Nia form Asabotsy (Saturday). 6. Nia and Masina form Alahady (Sunday).

7. Masina and Fàhasivy form Alatsinainy (Monday).

That is to say, if a combination of the two rubrics Trano and Lålana in the sikidy you have erected gives you a figure which is like Talé (which represents the man in question), the man is in danger of being taken ill on Tuesday. The procedure is the same with the other days.

The dangerous day being found out, the mpisikidy takes a piece of Anandrofolsy (an herb), puts it into a bottle (tavoàra), pours water on it till it is brimful, puts a cane (Bàraràta) into it, by which of course some water must overflow; and this he takes and rubs or brushes it seven times on the man from head to foot, saying, "His day is conquered by him, his vintana is conquered (overcome, leony') by him, and does not overcome him. His day and his vintana shall not govern him (‘mìtondra azy'), but he shall govern them."

As this and the preceding section is neither called vintana or san-andro, although it is in reality only a peculiar form of the last, and as san-andro is itself only a form of vintana (planetary vintana, I ought rather to have called the whole of this chapter Vintana, and subdivided it into ‘Zodiacal and Lunary Vintana,' and 'Planetary Vintana.' To the first would belong everything depending on the Zodiac and the Moon-stations and which is connected with the month-days; to the second all that depends on the seven planets and is closely connected with the week.

The question occurred to me: What is to be done, if a day is a lucky one as to its place in the month, but unlucky according to its position in the week? But as the first class chiefly referred to birth and business, the second to burial, offerings and diseases, I suppose these sly diviners managed to avoid a collisio officiorum.

There are many other points in sikidy and vintana which I have been obliged to pass over in silence, or only slightly touch upon. But it is time that I should bring this article to a close.

The sikidy and vintana was once the most tremendous power in this island; let us thank God that its spell is broken, and its influence passing away.

L. DAHLE.

NOTE.-Let me take this opportunity of correcting an error in the first of these papers (ANNUAL X., 1886, p. 228) I expressed some doubt as to the phrase mamo-hefa. A native has pointed out to me that it certainly should be mamòha éfa, 'to revive again a past (evil),' e.g. a disease.-L.D.

THE

VOLCANIC LAKE OF TRITRIVA :

ITS PHYSICAL FEATURES AND LEGENDARY HISTORY.

TH

HE great island of Madagascar is not at present one of those regions of the earth where volcanic disturbances occur; but there is ample evidence from the numerous extinct craters found in various parts of the island that at a very recent period, geologically considered-- possibly even within the occupation of the country by its present inhabitants-it was the theatre of very extensive outbursts of subterranean energy. The whole island has not yet been examined with sufficient minuteness to determine the exact extent of these old volcanoes, but they have been observed from near the south-east coast in S. Lat. 23°, and in various parts of the centre of the island up to the north-west and extreme north, a distance of 680 miles; and probably a more careful survey would reveal other links connecting more closely what is, as at present known, only a series of isolated groups of extinct craters. In the central provinces of Madagascar there are two large clusters of old volcanic cones and vents: one of them in about the same latitude as the Capital (19° S.), but from 50 to 70 miles away to the west of it, in the neighbourhood of Lake Itàsy ;

the other in the district called Vàkinankàratra, situated about 80 miles to the south-south-west of Antananarivo, and south-west of the great central mountain mass of Ankaratra.

This second volcanic region stretches from 20 to 30 miles from Antsírabè away west to Bètàfo and beyond it, and contains numerous and prominent extinct craters, such as Ivòko, Iatsífitra, Vòhitra, Tritríva, and many others, some of which have been described by the graphic pen of the late Dr. Mullens in his Twelve Months in Madagascar (pp. 214-219). The doctor says that he counted in this southern group about 60 cones and craters.*

There are also many hot springs in this Vakinankaratra region, the most noted of which are those at Antsirabe. At this place one of the chief springs is largely charged with lime, which has formed an extensive deposit all over a small level valley sunk some 20 feet below the general surface of the plain around the village. For a long time past this place has furnished almost all the lime used for building in the Capital and the central province of Imèrina. Besides the deposit over the floor of the valley, there is also a compact ridge shaped mass of lime accretion, 70 feet long by 18 to 20 feet wide, and about 15 or 16 feet high. This has all been deposited by the spring, which kept a passage through the lime to the top. Within the last eight or ten years, however, the spring has been tapped by a shaft of no great depth a few yards to the north, over which a large and commodious bath-house has been erected by the Norwegian Lutheran Mission; and here many visitors come to bathe in the hot mineral water, which has been found very beneficial in rheumatic and other complaints. A little distance to the south-west is another spring, not however hot, but only milk-warm, the water of which is drunk by those who bathe in the other spring. This water has been shown to be, in chemical constituents, almost identical with the famous Vichy water of France. All over the valley the water oozes up in various places; and about half a mile further north are several other springs, somewhat hotter than that just described, to which the natives largely resort for curative bathing.

During the excavations for the foundations of the bath-house, the skeletons of several examples of an extinct species of Hippopotamus were discovered, the crania and tusks being in very perfect preservation. Some of these are now in the Museum at Berlin; but the finest specimen was sent to the Museum of the University of Christiania in Norway. This Madagascar Hippopotamus was a smaller species than that now living in Africa, and is probably nearly allied to, if not identical with, another Hippopotamus (H. Lemerleï), of which remains were found in 1868 by M. Grandidier in the plains of the south-west coast (see p. 438, ante). I was informed by the people that wherever in these valleys the black mud is dug into for a depth of three or four feet, bones are sure to be met with. Probably a series of excavations would reveal the remains of many animals, birds and reptiles formerly inhabiting Madagascar. From the internal structure of the teeth and bones of the Hippopotami discovered at Antsirabe, traces of the gelatine being still visible, it is evident that the animals had been living at a comparatively recent period. There have been occasional * See ANNUAL IX., pp. 66-74,

+ Ibid., p. 74, for analysis of water.

vague reports of the existence of some large animal in the southern parts of the island; possibly the Hippopotamus is not yet absolutely extinct there; and perhaps the half-mythical stories of the Songòmby, Tòkandia, Làlomèna and other strange creatures current among the Malagasy are traditions of the period when these huge pachyderms were still to be seen in the lakes and streams and marshes of Madagascar.

Within a few miles of Antsirabe are two crater-lakes. The nearer and larger of these is called Andraikíba, which lies distant about four miles due west. This is a beautiful sheet of water, blue as the heavens in colour, in shape an irregular square, but curving round to the north-west, where it shallows into a marsh, which is finally absorbed in rice-fields. The lake is said to be of profound depth, but the hills surrounding it are not very lofty, rising only about 200 feet above the surface of the water, from which they ascend steeply. Fish and water-fowl and crocodiles also are said to be very abundant in and on its waters.

But the most interesting natural curiosity to be seen in the neighbourhood of Antsirabe is the crater-lake of Tritriva. This is situated about ten miles to the south-west, a pleasant ride of two hours by palanquin. Travelling at first in a westerly direction, the road then turns more to the south-west, and skirts the southern foot of the old volcano of Vohitra (already mentioned). Passing about a mile or so south of the high ground round the southern shores of the Andraikiba lake, the road gradually ascends to a higher level of country, so that in about an hour and a half's time we are nearly as high as the top of Vohitra-probably about 500 feet. Reaching a ridge between two prominent hills, we catch our first sight of Tritriva, now from two to three miles distant in front of us. From this point it shows very distinctly as an oval-shaped hill, its longest axis lying north and south, and with a great depression in its centre; the north-eastern edge of the crater wall being the lowest part of it, from which point it gradually rises southwards and westwards, the western edge being at the centre from two to three times the height of the eastern side. To the north are two much smaller cup-like hills, looking as if the volcanic forces, after the main crater had been formed, had become weaker and so been unable to discharge any longer by the old vent, and had therefore formed two newer outlets at a lower level.

Descending a little from the ridge just mentioned, we cross a valley with a good many scattered hamlets, and in less than half an hour reach the foot of the hill. A few minutes' pull up a tolerably easy slope, perhaps 200 feet in height, brings us to the top, at the lowest part of the crater edge; and on reaching the ridge the crater of the old volcano and its lake is before us, or, rather, below us. It is certainly an extraordinary scene, and unique of its kind. The inner sides of the crater dip down very steeply on all sides to a deep gulf, and here, sharply defined by perpendicular cliffs all round it, except just at the southern point, is a rather weird-looking dark-green lake far below us, the water surface being probably from 200 to 300 feet lower than the point we are standing upon, and consequently below the level of the surrounding country. The lake, exactly shut in by the cliffs of the crater surrounding it, is not blue in colour, like Andraikiba, although under a bright and cloudless sky, but a deep and somewhat blackish green. It must look, one would suppose, like ink under a stormy sky or in the shadows of evening.