stantive verb es-mi (the original form), es-si, es-ti (still existing in this form in the Lithuanian language), of the old Latin verb es-um, es, es-t, es-umus, es-tis, es-unt, es-to, es-se, and with a slight variation of the Sanscrit as-mi, &c. With the same form of the Latin we may connect es-t, he eats, es-se, to eat, es-ca, es-culentus, &c., and the German ess-en, to eat. After the word had thus been stripped of its initial consonant, the short vowel also was apt to disappear, at least in the longer forms. Thus from the old Latin forms esum, esunt, esim, &c., there arose the shorter forms sum, sunt, sim, &c.; prae-es-ens, ab-es-ens, were reduced to praesens, absens; and in German we find sein, to be, sind, they are, in place of es-ein, es-ind.

In the second place, the consonant s interchanges with the letter r [see AUSONES], so that were exists by the side of was, and art, are, with is. Thus in the Latin, too, we have er-am, er-o, where more regular forms would have been es-am, es-o, or rather es-so. Again, the same letter s is interchangeable with the dentals t, d. Hence, while the Germans have ess-en, Ich ass, the English express the same notions by to eat, I ate; and the Latin tongue uses indifferently ed-it or es-t, he eats, ed-ere or es-se, to eat.

The form be is evidently the parent of the German bin, I am, bist, thou art, and of the English be-ing and be-en. With the short vowel changed, it appears in the Lithuanian bu, as bu-ti, to be, buwau, I have been; and as the latter language is closely allied to the Greek, and other IndoGermanic tongues, we cannot be surprised at the Greek form fu-o, I beget, &c. pe-fu-ca, and e-fu-n, I am, &c.; whence also the Latin fu-vi or fu-i, fu-am, fu-turus, &c. That these forms are all related among themselves is generally allowed; but the question now proposed is, whether they are not also radically connected with the root wes. If it could be shown that the root be ever existed with an s at | the end, it would no longer be thought a violent step to suppose a connection between bes and wes, more especially when we find the b already half way towards a w in fui. In foetus, foecundus, &c., pronounced probably fuetus, &c., we have a still nearer approach to the digamma. Now a strong presumption that the root be had a sibilant, arises from the old German form birumes, we are, compared with warumes, we were, in the same language (see Grimm). In these words the suffix, which denotes the plural pronoun, cannot claim more than the four letters umes, thus agreeing very precisely with the Greek suffix omes, the Latin umus, and the Lithuanian ame of the same power. The radical parts then are bir and war; and as we know the latter to be connected with the form was, there is no slight suspicion that bir implies an early form, bis. If the Greeks lost the s in many of their forms derived from the short root es, as they did, and if we ourselves have dropped it from am, we can scarcely be surprised at its disappearance from the longer form bes or bis. The notion that the roots bes and wes are connected, is confirmed again by the other forms in these languages, which represent the idea of eating. In Greek, we find bo-sco, bo-tos, bo-ra, in Latin pasco, pascor, as well as vescor. The root pas is another instance of the arbitrary retention or omission of the sibilant, as we have pas-tor, pas-tus with the sibilant, pa-bulum without it.

The use of this auxiliary in the passive, both in ancient and modern languages, is familiar to all; but it has been less carefully observed, that it is likewise employed in the perfect tenses of the active voice, at least in the Latin language. Amav-eram, amav-ero, amav-issem, amuv-isse, evidently contain the forms eram, ero, essem, esse; and in the perfect subjunctive, an older form, amavesim, may be inferred from the three existing forms amassim, amaverim, amarim; and in amav-esim we see the full form esim which preceded sim. After the verb to be, the next in importance among the auxiliaries is the verb habe-re, Latin, to have; in German, hab-en. Like the preceding verb, this also has undergone great corruptions. In the English hast, has, had, the main consonant has already disappeared. While in the Italian ho, from the Latin habeo, we find nothing of the root but the aspirate, and even that is often omitted, so that we should doubt the connection between the words but for the first and second persons plural. But as we shall have further occasion for the forms of this verb in the Roman languages of Europe, we will place here the present tenses. Latin, habeo, habes, habet; habemus, habetis, habent. Italian, ho, hai, ha; abbiamo, avete, hanno. Spanish, he, has, ha; habémos or hémos, habéis, han. French, ai, a; avons, avez, ont.

as,

But

The use of the verb to have in the formation of the perfects, so universal in the modern languages derived from Latin, may be occasionally seen in the parent language also, where such phrases as furem constrictum habes, fures constrictos habeo, differ but slightly in meaning from furem constrinxi, &c.; and there was the greater necessity for adopting a new formation, as the Latin perfect unites two tenses in itself, viz., the aorist and the present-perfect. It will be seen, too, from the examples which we have given, why, in the derived tongues, the participle in some cases agrees with the accusative; as je les ai tués. the use of habeo as an auxiliary is not confined to the perfect tenses. In connexion with the infinitive it forms a convenient periphrasis for a future. From the Italian infinitive sentir, we have a future sentir-o, -ai, -a, -emo, -ete, -anno, the first and second persons plural, now they are used as suffixes, being reduced as completely as the rest In the Spanish verb hablar the future is hablar-é, -as, -á, -émos, -éis, -án; and in the French, from sentir there is formed sentir-ai, as, -a, -ons, -ez, -ont. In the tense called generally the conditional, the infinitive is again employed. The Italians unite with it their perfect tense of to have, derived from habui, viz., ebbi, avesti, ebbe, avemmo, aveste, ebbero; and their conditional is sentir-ei, -esti, -ebbe, -emmo, -este, -ebbero. On the other hand the French employ avois, which may be proved to have been derived from the Latin imperfect habebam (see Raynouard); but as avons, avez, of the present dropped their radical letters ar when attached as suffixes to the future, so also avois, &c., throughout lose the same letters in forming the conditional, thus, sentir-ois, -ois, -oit, -ions, -iez, -oient. The Spanish language, in like manner, employs the imperfect habia, habias, había, habíamos, habíais, habian, derived also from habebam, &c.; and thus, with the same suppression of the two first letters, the conditional of hablar is hablar-ia, -ías, -ía, -íamos, -iais, -ian. This view of the formation of the futures is of service in explaining the apparent irregularities so often found in those tenses, which moreover generally extend to the infinitive.

Many other verbs of the Latin language have become auxiliaries in the derived languages. 1. Vado, Lat. I go, is employed thus by the Italians, as io vo faciendo, I am doing, and in French for a future, je vais parler, I am going to speak. 2. Venio, Lat: I come, in Italian as an equivalent for the verb to be: egli vien riputato, he is considered; in French to denote an action just passed: il vient de trouver, he has just found. 3. Ambula-re, to walk, (corrupted into the Italian andare and the French aller,) is used in the former language thus, andra rovinato, he will be ruined, and in the French, il alloit diner, he was going to dine. 4. Sta-re, to stand, in Italian sono stato, I have been, sta scrivendo, he is writing; and the French étois (formerly estois) is a corruption from stabam, precisely as aimois from amabam. The Spaniards, besides several of the auxiliaries here mentioned, use tener, derived from the Latin tene-re, to hold, but not exactly as an auxiliary verb: and besides ser, to be, they have estár, to be, from the Latin stare. In the Teutonic languages the auxiliary verbs are very numerous, and our own language contains nearly the whole of them: 1. may, might, are the present and perfect of the same defective verb. In the German we find an infinitive of this verb, mög-en, as well as the forms mag, and mochte; 2. can and could correspond to the German kann and konnte from the infinitive könn-en; 3. will and would to the German will and wollte from woll-en; 4. shall and should to soll and sollte from sollen.

But though the German auxiliaries correspond with the English as to their having a common origin, they have a use which is not quite the same. In general, possibility is expressed by können, dürfen (the English dare, durst), mögen, and necessity by müssen (the English must), sollen, wollen; lassen (the English let) implies necessity as well as possibility. (Becker's German Grammar, p. 65.) The German word haben, like the corresponding English have, and the German werden, when used alone, are notional verbs, or verbs expressing distinct notion and not a mere relation: thus we can say, er wird reich, he becomes rich; but in the expressions ich werde kommen, I will come, die frage wird von ihm beantwortet, the question is answered by him, the verb werden is used as an auxiliary for the future tense and the passive voice respectively.

In the antient Greek language it has not been observed till of late years, nor, indeed, is it yet universally admitted,

that the verbs to have and to wish (Exw and Okλw) are often used as auxiliary verbs. That such, however, is the fact, will not be disputed by those who are the best judges. The forms of the auxiliary verbs in the modern Greek language are a confirmation of this opinion.

Those who wish to compare the forms of the Lithuanian language with those of the cognate tongues, may consult Anfangs-gründe einer Littauischen Sprachlehre, &c., Von C. G. Mielcke, Königsberg, 1800.

AUXONNE, or AUSSONNE, a fortified town in France, on the left bank of the Saône, and on the road from Paris through Dijon to Geneva; 201 miles S.E. of Paris and 18 S.E. of Dijon. It is in the department of Côte d'Or.

The foundation of this city is fixed by some about the year 400 of the Christian æra, but nothing certain appears to be known of it. It was first part of the county of Burgundy, and then was made the capital of the county of Auxonne, which came by exchange into the hands of the dukes of Burgundy, but was not united with their duchy. Upon the seizure of the possessions of the dukes of Burgundy by Louis XI., it fell into the power of that prince. By the treaty of Madrid, it was ceded by Francis I. of France to the emperor Charles V.; but the inhabitants refused to submit to the emperor, and obliged his general, Lannoy, who in 1526 besieged the town for nine months, to raise the siege. Before the kings of France acquired possession of the Franche Comté, Auxonne was one of the frontier towns of Burgundy, and was defended by a castle flanked by six large towers, erected by Louis XI., Charles VIII., and Louis XII. In 1673, the town was fortified by Vauban. The streets are straight, and the fortifications serve as pleasant promenades. There is a fine bridge over the Saône; and at the end of the bridge, on the side next the town, is a causeway of 2350 paces in length (nearly 24 English miles), having twenty-three arches for allowing passage to the water in case of inundation. This causeway was constructed or improved by Margaret of Bavaria, duchess of Burgundy, in 1405. There are barracks, an arsenal, a school of artillery, a cannon foundry, and powder mills: also, a library and a high school. There were several religious houses before the Revolution, and a poor and ill-built hospital. The town formerly consisted of two parishes, which are now incorporated into one.

The trade of Auxonne consists chiefly in wine, grain, and wood; as well as cloth and serges, which are sent to Lyon. In return, groceries, silk, and the wines of Mâcon, are received. These wines are sold again in Lorraine and Franche Comté. There are several working goldsmiths in the town. The number of inhabitants was, in 1832, about 5,000.

There are, in the neighbourhood of Auxonne, quarries of stone of various colours-blue, red, yellow, and purple. Some species of turquoises and fossil corals are found in these quarries. There is also pretty good marble. The district of Auxonnois is low and marshy, watered by many brooks, and affording abundance of good pasturage. (Encyclopédie Méthodique; Dictionnaire Universel de la France; Dupin; Malte Brun; Expilly.)

AUZOUT, ADRIEN, was a native of France, but neither the place nor time of his birth is known. He had established his reputation as an astronomer in 1666, and was one of the original members of the Academy of Sciences, founded in that year. He died in 1691, according to Delambre and others; in 1693, at Rome, according to Montucla, who cites the records of the Academy. The collection hereafter noticed, published in 1693, speaks of him as living (p. 340). Auzout is celebrated as having, in conjunction with Picard, applied the telescope to the mural quadrant. This rests on an admission of Picard to Lahire (Montucla, ii. 569), asserted by the latter; but there is no mention of it in Picard's book On the Figure of the Earth. Auzout also made an independent invention and application to the telescope of the moveable wire micrometer, on which he published a treatise in 1667. Picard assisted him in perfecting this instrument. Huyghens has been frequently stated as an inventor of this micrometer, but his instrument is different from, and inferior in principle to, that of Auzout. (Delambre, Ast. Mod., Disc. Prelim. p. 47.) The prior invention of Gascoyne (ASTRONOMY, p. 534) is admitted, and was brought forward by Hooke and others of the Royal Society, in opposition to the invention of Auzout. The French have always retorted, with just severity, that the invention of their countrymen must needs be original, since the English themselves did not know what Gascoyne had

done, till Auzout communicated his own micrometer to the Royal Society; thus allowing a method of first-rate importance to astronomy to lie idle, till national feeling, and not love of science, ransacked their own records.

Auzout published observations and calculations of the comet of 1664, and the presentation of his results to Louis XIV. is said to have given that prince the first idea of founding an observatory at Paris. He also made a laborious comparison of the weights and measures of France and other countries, which is to be found, together with his own account of his micrometer, in the folio collection of Memoirs of the Academy, entitled Divers Ouvrages de Mathématique et Physique, Paris, 1693. Among other results of the micrometer, he observed and measured the diurnal variation of the moon's diameter, first explained by Kepler. He was engaged in several discussions with Hooke, which more concern the biography of the latter.

Besides the preceding works, we have left of Auzout a letter on some new observations of Jupiter and Saturn, Paris, 1664; and a letter to the Abbé Charles on a collection of observations published by Campani, Paris, 1665.

Auzout was a good optician and maker of telescopes; and when it is added that he never enjoyed even tolerable health, he must be considered as having done not a little for astronomy.

AVA. [See BIRMAN EMPIRE.]

AVAL. [See BAHREIN ISLANDS.]

AVALANCHES are the most dangerous and terrible phenomena to which the valleys embosomed between high, snow-topped mountain ranges are exposed. They are especially frequent in the Alps, owing to the steepness of their declivities, but they are also known in other mountain regions, as in the Pyrenees and in Norway. They originate in the higher region of the mountains, when the accumulation of snow becomes so great that the inclined plane on which the mass rests cannot any longer support it. It then is pushed down the declivity by its own weight, and precipitated into the subjacent valley, where it often destroys forests and villages, buries men and cattle, and sometimes fills up the rivers and stops their course. Besides what is covered with the masses of snow, persons are often killed and houses overthrown by the sudden compression of the air, caused by the incredible velocity with which these enormous masses descend.

Four different kinds of avalanches may be distinguished: drift avalanches, rolling avalanches, sliding avalanches, and glacier or ice avalanches, of which the first commonly take place in the early part of the winter, the second and third at the end of winter and in spring, and the last only in summer.

The drift or loose snow avalanches (called, in Switzerland, staub-lauinen) take place when heavy snow has fallen in the upper region of the mountains during a still calm, and this accumulated mass, before it acquires consistency, is put in motion by a strong wind. The snow is driven from one acclivity to another, and so enormously increased in its progress, that it brings down an incredible volume of loose snow, which often covers great part of a valley. The damage caused by these avalanches is, however, generally not very great, because most of the objects covered by them may be freed from the snow without having sustained great damage; but they often produce such a compression of the air that houses are overturned, and men and cattle suffocated.

The rolling avalanches are much more dangerous and destructive. These take place when, after a thaw, the snow becomes clammy, and the single grains or flocks stick to one another, so as to unite into large hard pieces which commonly take the form of balls. Such a ball, moved by its own weight, begins to descend the inclined plane, and all the snow it meets in its course downwards sticks firmly to it, This snow-mass, increasing rapidly in its progress, and de scending with great velocity, covers, destroys, or carries away everything that opposes its course-trees, forests, houses, and rocks. This is the most destructive of the avalanches, and causes great loss of life and property. In the year 1749, the whole village of Rueras, in the valley of Tawich, in the canton of the Grisons, was covered, and at the same time removed from its site, by an avalanche of this description; but this change, which happened in the night time, was effected without the least noise, so that the inhabitants were not aware of it, and on awaking in the morning could not conceive why it did not grow day. A hundred persons were dug out of the snow, sixty of whom were still alive, the interstices between the snow containing sufficient air to sup

VOL. III.-Y

Avallon, and from thence sent in curiously-constructed rafts down the Cousin into the Cure; by this again into the Yonne, and then by the Seine to Paris. Casks, mustard, woollen cloth, and paper are among the manufactures of the town and neighbourhood, and a trade is carried on in corn, wine, cattle, and in coals. The inhabitants are above 5000. Avallon is the capital of an arrondissement or subprefecture, and has a tribunal de commerce, or court of reference for settling commercial disputes, a high-school (collège), and an agricultural society. There are a theatre, public baths, and an hospital. Before the revolution there were four religious houses, and the church was collegiate. Near Avallon may be traced some remains of the Roman road formed by Agrippa, son-in-law of Augustus, from Lugdunum (Lyon) to Gesoriacum (Boulogne).

The neighbourhood of Avallon consists, on one side, of meadow land and marshes, and on the other of arable land, fertile in corn, and producing wines of good quality. The arrondissement contains 456 square miles, or 291,840 acres. Its population is about 47,000. (Expilly, Dictionnaire Géographique, &c., des Gaules et de la France; Millin, Voyage dans les Departements du Midi de la France; Letters from France, by John M. Cobbett; Malte Brun.) AVANTURINE, a variety of quartz, remarkable for the brilliancy with which it reflects light, the effect being in general produced by fine points of mica imbedded within the crystalline mass. From this circumstance it is sometimes employed in jewelry, but it is of little value.

8. The most celebrated of the Avataras of Vishnu is his appearance in the human form of Krishna, in which he is supposed to have been wholly and completely incarnate, whereas the other Avataras are only considered as emanations from his being. Krishna assisted the family of the Pândavas in their war with the Kurus, and through them relieved the earth from the wicked men who oppressed it. The history of this conflict is told at length in the Mahâbhârata, another great epic poem in Sanskrit.

9. Buddha is, by the followers of the Brahmanical religion, considered as a delusive incarnation of Vishnu, assumed by him in order to induce the Asuras to abandon the sacred ordinances of the Vedas, by which they lost their strength and supremacy.

10. Kalki is the name of an Avatâra in which Vishnu will appear at the end of the Kaliyuga, or present age of the world, to destroy all vice and wickedness, and to restore the world to virtue and purity.

We cannot enumerate the Avatâras of the inferior deities, in which the mythology of the Hindus abounds. We do not remember ever to have heard of any of Brahmâ or Siva, the two supreme deities who, with Vishnu, constitute the Trimurti. In the seventh volume of the Asiatic Researches (Calcutta, 1801) may be seen an account given by Captain Edward Moor of an incarnation of Ganesa, or Ganapati, which had, since the year A.D. 1640, become hereditary in the family of Mooraba Gosain, a Brahman at Punah. Mrs. Graham (now Mrs. Callcott), who in 1809 visited this living Avatara, which was then a child, has given an interesting notice of it in her journal.

mould here covers a continuous chalk substratum, which presents a clean, dry, grassy surface. The immediate site chosen for the grand circular temple is a flat area of ground, having an irregular range of gentle hills to the east, running north and south, a rising tract of land to the south, a level country of some miles in extent to the north, some undulating and rather high hills to the west, and extending thence to the east. A small brook, or rivulet, called the Kennet, a tributary to the Thames, has its source a short distance north of the temple. The geological characteristics of the district probably occasioned its being chosen for the erection of a temple which was to be the chief edifice of the sort in the island. As Stukeley remarks, this might have been regarded as the grand national cathedral, while the smaller circles, in different parts of the island, might be compared to the parish, or village churches. On the surface of the ground, both in the neighbouring valleys and on the high lands, are numerous large masses of stone. There is still a great number of detached oolitic sandstones of various sizes, known by the name of the Grey Wethers, lying near the Bath road, in the neighbourhood of Avebury. From amongst these stones scattered about the neighbourhood, the builders or rather makers of the temple selected such as seemed best adapted to their rude design.

No less than 650 blocks were brought together and placed in circles and rows. These stones were of various dimensions, measuring from five to twenty feet in height above the ground, and from three to twelve feet in width and thickness. One hundred were raised on end, and placed in a circular form, around a flat and nearly circular area of about fourteen hundred feet in diameter; and these stones were bounded by a deep ditch and lofty bank, which enclosed the whole work, except at two places, where open

(See the articles MANU, RAMA, KRISHNA, BUDDHA; Bohlen, Das alte Indien, vol. i. pp. 213-234; Vans Kennedy Researches into the Nature and Affinity of Antient and Hindu Mythology, London, 1831, 4to.) AVEBURY, ABURY, ABIRY, the name of a villageings were left for entrances to the temple. The bank or and parish in Wiltshire, England, is remarkable as the site of what was once the largest and most interesting Celtic or Druidical temple in Europe. The origin of the name is uncertain: the last part, bury, a borough, or fortified place, appears to be a Saxon word, and if so, Avebury is not the original name of the place. We shall endeavour to furnish the reader with an account of this great work, as it is presumed to have been when in a perfect state; and also with the opinions of some eminent antiquaries respecting its original destination. For this purpose we must refer to the descriptions, drawings, and printed accounts which Dr. Stukeley has left in his interesting work, entitled Abury, a Temple of the British Druids, fol. 1743. Though the volume bears this date, it appears that the doctor surveyed the place, and made numerous drawings of it, and of the various objects in the vicinity, during the years 1720-24. It is due to that learned and zealous, but rather credulous antiquary to say, that his delineations have every appearance of general accuracy, and that his descriptions are sufficiently perspicuous to be understood.

mound at present is broken down in four places, but there seem to have been originally only two openings corresponding to the two great avenues which will be described hereafter. The inner slope of the bank measured eighty feet, and its whole extent, or circumference, at the top was, according to Sir Richard Colt Hoare, 4412 feet: the area within the bank or mound is somewhat more than twentyeight acres. About half way up the inner slope was a sort of terrace walk, apparently adapted for spectators. Dr. Stukeley conjectures that there was a second circular arrangement of upright stones at a short distance within the other circle; and he founds his opinion upon the fact of there being one large stone in a position which does not come into any other component circle of the temple. As shown in the accompanying diagram, No. 1, and view, No. 3, there were two other small temples within the periphery of the great circle. One was a double circle of upright stones, with a single stone raised near the centre, which Stukeley calls the ambive, or obelisk: this small temple consisted of forty-three stones, and is referred to in Before Stukeley's publications, very little was known of No. 1 at c. Another temple, of forty-five stones, some of Avebury; neither Camden, Leland, nor any other topo- which are still standing and of immense size, was placed a grapher or antiquary seems to have published any account little north of the former, and consisted also of two concenof it. Aubrey, a native of Wiltshire, and an ardent lover of tric circles, enclosing a group of three tall stones, called the antiquarian subjects, had visited Avebury in 1648. In the cove. These were the component parts and the general year 1663, he was commanded by King Charles II. to write design and arrangement of the triple temple, as it may some account of this remarkable monument, the monarch be called; but there were two members, or connecting being then on his way to Bath, and having examined parts, which belonged, and gave a peculiarity to this work, the whole in company with Aubrey and Dr. Charlton, who distinguishing it from all other Celtic temples. These afterwards published An Account of Stonehenge. These were avenues of approach, consisting of double rows, or antiquities, says Aubrey in his MS. treatise, are so ex- lines of upright stones, which branched off from the cenceedingly old that no books do reach them. I can affirm tral work, each to the extent of more than a mile. One of that I have brought this temple from utter darkness into them branched off from the outer circle, to the south, turna thin mist, and have gone farther in this essay than any ing, near its extremity, to the south-east, where it terbefore me. Though Aubrey's account is very imperfect, minated in two circular or rather elliptical ranges of upright and was never completed, and though his sketches are stones (see No. 2, e). According to Stukeley, this avenue even more defective than his descriptions, yet as they are was formed by two hundred stones, being finished at its the first records of the place, and contain some useful eastern extremity with fifty-eight stones. The width of the facts and evidence of the state of the temple seventy-six avenue varied from fifty-six to thirty-five feet between the years before Dr. Stukeley commenced his survey, they are stones, which were, on an average, eighty-six feet apart from valuable and interesting to the antiquary. [See AUBREY.] each other in their linear direction. The outer oval of the The accompanying plans show the general form and terminating temple (e) to the south-east, on an eminence arrangement of the temple, with the avenues of approach. called Overton Hill, or the Hakpen, measured about 146 When the aborigines of the island fixed on this site for their feet in diameter; the inner oval was forty-five feet across. great temple, we may reasonably conclude that it was in a The western avenue (c) extended about one mile and a state of nature, and, like the general character of the Marl-half, and consisted of 203 stones; its extremity ended in a borough Downs and Salisbury Plain, was without either point, or with a single stone. These avenues, or grand forest-trees or underwood of any kind. A thin stratum of approaches to the temple, were not arranged in straight,

[No. 2.-Plan, or Map of the whole Temple, with its two avenues, c and d; the temple, a; a of the same from 1 to 2-i. e. from east to west. The plan, small temple, e; Silbury Hill, f; high ground, g; a line of road, or British track-way, h; the though on a small scale, shows the relative proportions and course of the river Kennet, i; line of Roman road from Bath to London, k; barrows; sites arrangements of the lofty bank, or vallum, e; the ditch, or of villages, 1.]

[No. 3.-A bird's eye View of the Temple, from the north, looking south, and intended to display the circumvallated bank, 1; the two inner, or small temples, 2 and 3; the course of the Kennet river, 4; the western avenue, 5; the southern avenue, 6; the situation of Silbury Hill, 7; a large barrow, called by Stukeley the Druids' barrow, 8; a cromlech, surrounded by a circle of small stones, 9.]